Docling Serve API#

The docling-serve-api module defines the core, framework-agnostic Java API used to communicate with a Docling Serve backend. It provides the request/response model and the main DoclingServeApi interface. You can use any implementation of this interface to talk to a running Docling Serve instance.

The base Java version is 17. This module has no other required dependencies, although it is compatible with both Jackson 2.x and 3.x.

If you need a ready-to-use HTTP implementation, see the reference client: - Docling Serve Client: docling-serve-client

When to use this module#

You want a stable, typed Java model for Docling Serve requests/responses without committing to a specific HTTP stack.
You plan to create your own client implementation (e.g., different HTTP lib, reactive runtime) but still rely on the shared API model.
You want to write code that is portable across different client implementations.

Installation#

Add the API dependency to your project.

GradleMaven

dependencies {
  implementation("ai.docling:docling-serve-api:0.6.2")
}

<dependency>
  <groupId>ai.docling</groupId>
  <artifactId>docling-serve-api</artifactId>
  <version>0.6.2</version>
</dependency>

Note: The API module does not perform network I/O by itself. To call a live service, combine it with an implementation such as docling-serve-client.

Quick start#

Below is a minimal example using the reference client to create an implementation of DoclingServeApi, build a conversion request, and retrieve Markdown output. The request/response types all come from docling-serve-api.

import java.net.URI;
import ai.docling.serve.api.DoclingServeApi;
import ai.docling.serve.api.convert.request.ConvertDocumentRequest;
import ai.docling.serve.api.convert.request.options.ConvertDocumentOptions;
import ai.docling.serve.api.convert.request.options.OutputFormat;
import ai.docling.serve.api.convert.request.source.HttpSource;
import ai.docling.serve.api.convert.request.target.InBodyTarget;
import ai.docling.serve.api.convert.response.InBodyConvertDocumentResponse;

DoclingServeApi api = DoclingServeApi.builder()
    .baseUrl("http://localhost:8000") // your Docling Serve URL
    .logRequests() // log HTTP requests
    .logResponses() // log HTTP responses
    .prettyPrint() // pretty-print JSON requests/responses
    .apiKey("optionally your api key")
    .build();

ConvertDocumentRequest request = ConvertDocumentRequest.builder()
    .source(HttpSource.builder().url(URI.create("https://arxiv.org/pdf/2408.09869"))
        .build())
    .options(ConvertDocumentOptions.builder()
        .toFormat(OutputFormat.MARKDOWN) // request Markdown output
        .includeImages(true)
        .build())
    .target(InBodyTarget.builder().build()) // get results in the HTTP response body
    .build();

InBodyConvertDocumentResponse response = (InBodyConvertDocumentResponse) api.convertSource(request);
System.out.println(response.getDocument().getMarkdownContent());

Core concepts#

The `DoclingServeApi` interface#

Defined in ai.docling.serve.api.DoclingServeApi, this interface exposes many operations:

Any HTTP or non-HTTP implementation can implement this interface. The reference implementation is provided by the docling-serve-client module.

Extension Points#

The docling-serve-api module uses the Java Service Provider Interface to define an extension point for building/customizing instances of DoclingServeApi. An application (or downstream framework) can create an implementation of ai.docling.serve.api.spi.DoclingServeApiBuilderFactory and register it via the META-INF/services/ai.docling.serve.api.spi.DoclingServeApiBuilderFactory file, or providing an implementation using Java modules.

This is exactly what the docling-serve-client module does to provide its implementation. See module-info.java and DoclingServeClientBuilderFactory.java.

Requests: `ConvertDocumentRequest`#

Create a request via the builder:

ConvertDocumentRequest request = ConvertDocumentRequest.builder()
    .source(/* one of the supported sources */)
    .options(/* conversion options */)
    .target(/* optional delivery target */)
    .build();

Supported sources (ai.docling.serve.api.convert.request.source): - HttpSource — fetch content from a URL (optional custom headers) - FileSource — embed content as Base64 with a filename - S3Source — fetch content from an S3 bucket

Targets (ai.docling.serve.api.convert.request.target): - InBodyTarget — receive results directly in the API response body (default use case) - PresignedUrlTarget — the service uploads each output artifact to its configured object storage and returns time-limited presigned download URLs (requires docling-serve v1.22.0+) - PutTarget — the service uploads converted content via HTTP PUT to a specified URI - ZipTarget — receive a zipped result - S3Target — upload converted content to an S3 bucket

Options (ai.docling.serve.api.convert.request.options.ConvertDocumentOptions) let you control: - Input/output formats (e.g., fromFormats, toFormats) - OCR (e.g., doOcr, forceOcr, ocrEngine, ocrLang) - PDF processing (e.g., pdfBackend) - Tables (e.g., tableMode, tableCellMatching) - Pipelines and page ranges (e.g., pipeline, pageRange) - Timeouts and error behavior (e.g., documentTimeout, abortOnError) - Enrichments (code/formula/picture), image handling, scaling, page break placeholder - VLM pipeline hints

Explore the options package for the full list of knobs you can turn.

Responses: `InBodyConvertDocumentResponse`, `PreSignedUrlConvertDocumentResponse`, `PreSignedUrlConvertResponse`, `ZipArchiveConvertDocumentResponse`, and `DocumentResponse`#

InBodyConvertDocumentResponse contains the converted document (if any), errors, processing status, total processing_time, and detailed timings map.
PreSignedUrlConvertDocumentResponse contains processing statistics - total processing_time and conversion metrics num_converted, num_succeeded, num_partially_succeeded, num_failed.
PreSignedUrlConvertResponse is returned when using a PresignedUrlTarget. It contains per-document results in a documents list, where each DocumentArtifactItem carries the conversion status, any errors, and a list of ArtifactRef entries with presigned download URLs for each output format. It also includes the same aggregate conversion metrics as PreSignedUrlConvertDocumentResponse.
ZipArchiveConvertDocumentResponse contains file_name and an input stream for the archive.
DocumentResponse holds the actual content fields you requested, such as md_content (Markdown), html_content, text_content, and a json_content map. It also includes the filename and doctags_content when relevant.

Presigned URL target example#

When using a PresignedUrlTarget, the response contains per-document artifact download links:

import java.net.URI;
import ai.docling.serve.api.DoclingServeApi;
import ai.docling.serve.api.convert.request.ConvertDocumentRequest;
import ai.docling.serve.api.convert.request.source.HttpSource;
import ai.docling.serve.api.convert.request.target.PresignedUrlTarget;
import ai.docling.serve.api.convert.response.PreSignedUrlConvertResponse;

DoclingServeApi api = DoclingServeApi.builder()
    .baseUrl("http://localhost:8000")
    .build();

ConvertDocumentRequest request = ConvertDocumentRequest.builder()
    .source(HttpSource.builder().url(URI.create("https://arxiv.org/pdf/2408.09869")).build())
    .target(PresignedUrlTarget.builder().build())
    .build();

PreSignedUrlConvertResponse response = (PreSignedUrlConvertResponse) api.convertSource(request);
System.out.println("Converted: " + response.getNumSucceeded() + "/" + response.getNumConverted());

response.getDocuments().forEach(doc -> {
    System.out.println(doc.getFilename() + " -> " + doc.getStatus());
    doc.getArtifacts().forEach(artifact ->
        System.out.println("  " + artifact.getArtifactType() + ": " + artifact.getUri()));
});

Health checks#

You can ping the service to check readiness and basic status:

import ai.docling.serve.api.health.HealthCheckResponse;

HealthCheckResponse health = api.health();
System.out.println("Service status: " + health.getStatus());

Error handling#

Conversion may succeed partially (e.g., some pages) while returning warnings or errors. Always inspect InBodyConvertDocumentResponse#getErrors() and consider status:

InBodyConvertDocumentResponse response = (InBodyConvertDocumentResponse) api.convertSource(request);

if (response.getErrors() != null && !response.getErrors().isEmpty()) {
  response.getErrors().forEach(err ->
      System.err.println("Component: " + err.getComponentType() + 
          ", Module: " + err.getModuleName() + 
          ", Message: " + err.getErrorMessage()));
}

if (response.getDocument() != null && response.getDocument().getMarkdownContent() != null) {
  // Use the output
}

The exact transport-level exceptions (e.g., timeouts, connectivity) depend on the client implementation you use. The reference client throws standard Java exceptions for HTTP and I/O failures.

Validation errors#

In the case of a request validation error (i.e. docling-serve throws a 422 error), the docling-java API will throw an ai.docling.serve.api.validation.ValidationException which can be caught and inspected.

Logging and builders#

DoclingServeApi exposes a toBuilder() method so implementations can be duplicated and tweaked. Most client builders, including the reference client, also expose logRequests() and logResponses() for simple diagnostics:

DoclingServeApi newApi = api.toBuilder()
    .logRequests()
    .logResponses()
    .build();

Version compatibility#

The API is tested against all published versions of Docling Serve each week. Below are the latest run results:

Results for ghcr.io/docling-project/docling-serve as of 2026-08-24T04:37:21.526667782Z#

Here are the results:

Tag	Result	Details
v1.31.0	✅ SUCCESS	Click for run details
v1.30.0	✅ SUCCESS	Click for run details
v1.29.0	✅ SUCCESS	Click for run details
v1.28.0	✅ SUCCESS	Click for run details
v1.27.0	✅ SUCCESS	Click for run details
v1.26.0	✅ SUCCESS	Click for run details
v1.25.0	✅ SUCCESS	Click for run details
v1.24.0	✅ SUCCESS	Click for run details
v1.23.0	✅ SUCCESS	Click for run details
v1.22.1	✅ SUCCESS	Click for run details
v1.22.0	✅ SUCCESS	Click for run details
v1.21.0	✅ SUCCESS	Click for run details
v1.20.0	✅ SUCCESS	Click for run details
v1.19.0	✅ SUCCESS	Click for run details
v1.18.0	✅ SUCCESS	Click for run details
v1.17.0	✅ SUCCESS	Click for run details
v1.16.1	✅ SUCCESS	Click for run details
v1.15.0	✅ SUCCESS	Click for run details
v1.14.3	✅ SUCCESS	Click for run details
v1.14.2	✅ SUCCESS	Click for run details
v1.14.1	✅ SUCCESS	Click for run details
v1.14.0	✅ SUCCESS	Click for run details
v1.13.1	✅ SUCCESS	Click for run details
v1.13.0	✅ SUCCESS	Click for run details
v1.12.0	✅ SUCCESS	Click for run details
v1.11.0	✅ SUCCESS	Click for run details
v1.10.0	✅ SUCCESS	Click for run details
v1.9.0	✅ SUCCESS	Click for run details
v1.8.0	✅ SUCCESS	Click for run details
v1.7.2	✅ SUCCESS	Click for run details
v1.7.1	✅ SUCCESS	Click for run details
v1.7.0	✅ SUCCESS	Click for run details
v1.6.0	✅ SUCCESS	Click for run details
v1.5.1	✅ SUCCESS	Click for run details
v1.5.0	✅ SUCCESS	Click for run details
v1.4.1	✅ SUCCESS	Click for run details
v1.4.0	✅ SUCCESS	Click for run details
v1.3.1	✅ SUCCESS	Click for run details
v1.3.0	✅ SUCCESS	Click for run details
v1.2.2	✅ SUCCESS	Click for run details
v1.2.1	✅ SUCCESS	Click for run details
v1.2.0	✅ SUCCESS	Click for run details
v1.1.0	✅ SUCCESS	Click for run details
v1.0.1	✅ SUCCESS	Click for run details
v1.0.0	✅ SUCCESS	Click for run details

Details#

ghcr.io/docling-project/docling-serve:v1.31.0#

Click to expand

###### Message

Click to collapse

Tag v1.31.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:37:04 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:37:04 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:37:04 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:37:05 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:37:05 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:37:05 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:37:05 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:37:05 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:37:05 - docling_jobkit.connectors.plugins.defaults - Connector 'KafkaTargetProcessor' skipped — optional dependency not installed (No module named 'confluent_kafka'). Install the matching extra to enable it.
INFO:   04:37:05 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:37:05 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:37:05 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/health$,/livez$,/ready$,/healthz$,/readyz$)
INFO:   04:37:05 - uvicorn.error - Started server process [1]
INFO:   04:37:05 - uvicorn.error - Waiting for application startup.
INFO:   04:37:08 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:37:08 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:37:08 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:37:08 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:37:08 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:37:08 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 9d06178a9f1c8aeb9889d1e46440f3fa
INFO:   04:37:08 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:37:08 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:37:08 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-08-24 04:37:08.927188507 [W:onnxruntime:Default, device_discovery.cc:146 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:37:09 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:37:09,158 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:09,159 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:37:09,214 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:09,215 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:37:09,257 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:09,258 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:37:09 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:37:09 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:37:09 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 476/770 [00:00<00:00, 4746.98it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4804.17it/s]
INFO:   04:37:10 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:37:10 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:37:10 - uvicorn.error - Application startup complete.
INFO:   04:37:10 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:37:10 - docling_serve.app - Health check requested
INFO:   04:37:10 - uvicorn.access - 172.17.0.1:60414 - "GET /health HTTP/1.1" 200
INFO:   04:37:10 - docling_serve.app - Health check requested
INFO:   04:37:10 - uvicorn.access - 172.17.0.1:60420 - "GET /health HTTP/1.1" 200
INFO:   04:37:10 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:37:10 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:37:10 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:37:10 - docling_serve.app - [TENANT_ID] Task dc658ce0-e5aa-4e1d-aa92-2bb35dde963e created with tenant_id='default'
INFO:   04:37:10 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task dc658ce0-e5aa-4e1d-aa92-2bb35dde963e
INFO:   04:37:10 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:37:10 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a1d02772fb798c5f3e06aa82e0d32e7d
INFO:   04:37:10 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:37:10 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:37:10,760 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:10,760 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:37:10,812 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:10,812 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:37:10,854 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:10,854 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:37:10 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:37:10 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:37:10 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 481/770 [00:00<00:00, 4798.42it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4802.40it/s]
INFO:   04:37:11 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:37:11 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:37:11 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:37:11 - docling.document_converter - Going to convert document batch...
INFO:   04:37:11 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash a1d02772fb798c5f3e06aa82e0d32e7d
INFO:   04:37:11 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:37:12 - docling.document_converter - Finished converting document file in 0.40 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:641: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
WARNING:    04:37:12 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:37:12 - docling_jobkit.convert.results - Processed 1 docs in 0.47 seconds.
INFO:   04:37:12 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job dc658ce0-e5aa-4e1d-aa92-2bb35dde963e in 0.47 seconds
INFO:   04:37:12 - uvicorn.access - 172.17.0.1:60420 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:37:12 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:37:12 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:37:12 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:37:12 - docling_serve.app - [TENANT_ID] Task 0c5fa041-e811-411a-ad3b-533591cbfc11 created with tenant_id='default'
INFO:   04:37:12 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 0c5fa041-e811-411a-ad3b-533591cbfc11
INFO:   04:37:12 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:37:12 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a1d02772fb798c5f3e06aa82e0d32e7d
INFO:   04:37:12 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:37:12 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:37:12,781 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:12,781 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:37:12,834 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:12,834 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:37:12,881 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:12,881 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:37:12 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:37:12 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:37:12 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 478/770 [00:00<00:00, 4776.62it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4792.24it/s]
INFO:   04:37:13 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:37:13 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:37:13 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:37:13 - docling.document_converter - Going to convert document batch...
INFO:   04:37:13 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash a1d02772fb798c5f3e06aa82e0d32e7d
INFO:   04:37:13 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:37:13 - docling.document_converter - Finished converting document file in 0.16 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:641: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:37:13 - docling_jobkit.convert.results - Processed 1 docs in 0.16 seconds.
INFO:   04:37:13 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 0c5fa041-e811-411a-ad3b-533591cbfc11 in 0.16 seconds
INFO:   04:37:14 - uvicorn.access - 172.17.0.1:60420 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:37:14 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:37:14 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:37:14 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:37:14 - docling_serve.app - [TENANT_ID] Task f36274fd-42a8-45f9-98af-1e5bcfcb3f45 created with tenant_id='default'
INFO:   04:37:14 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task f36274fd-42a8-45f9-98af-1e5bcfcb3f45
INFO:   04:37:14 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:37:14 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 21c1cdafbd5b48f9b0cfcb806e9e9d57
INFO:   04:37:14 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:37:14 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:37:14,793 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:14,793 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:37:14,842 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:14,842 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:37:14,897 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:37:14,897 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:37:14 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:37:14 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:37:14 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 477/770 [00:00<00:00, 4762.95it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4817.50it/s]
INFO:   04:37:15 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:37:15 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:37:15 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:37:15 - docling.document_converter - Going to convert document batch...
INFO:   04:37:15 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 21c1cdafbd5b48f9b0cfcb806e9e9d57
INFO:   04:37:15 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:37:16 - docling.document_converter - Finished converting document file in 0.36 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:641: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:37:16 - docling_jobkit.convert.results - Processed 1 docs in 0.39 seconds.
INFO:   04:37:16 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job f36274fd-42a8-45f9-98af-1e5bcfcb3f45 in 0.39 seconds
INFO:   04:37:16 - uvicorn.access - 172.17.0.1:60420 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:37:16 - uvicorn.access - 172.17.0.1:60420 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:37:16 - uvicorn.access - 172.17.0.1:60420 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.30.0#

Click to expand

###### Message

Click to collapse

Tag v1.30.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:35:54 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:54 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:35:54 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:35:55 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:35:55 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:35:55 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:35:55 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:35:55 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:35:55 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:35:55 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:35:55 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/readyz$,/healthz$,/ready$,/health$,/livez$)
INFO:   04:35:55 - uvicorn.error - Started server process [1]
INFO:   04:35:55 - uvicorn.error - Waiting for application startup.
INFO:   04:35:58 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:59 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:35:59 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:59 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:35:59 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:59 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash e2db689ab35ca52a08b5b7fa25de8cc8
INFO:   04:35:59 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:59 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:35:59 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-08-24 04:35:59.672906055 [W:onnxruntime:Default, device_discovery.cc:134 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:35:59 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:35:59,979 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:35:59,980 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:36:00,082 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:00,083 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:36:00,174 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:00,175 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:36:00 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:36:00 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:36:00 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[transformers] `torch_dtype` is deprecated! Use `dtype` instead!

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  60%|██████    | 462/770 [00:00<00:00, 4607.84it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4320.50it/s]
INFO:   04:36:02 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:36:02 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:36:03 - uvicorn.error - Application startup complete.
INFO:   04:36:03 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:36:04 - docling_serve.app - Health check requested
INFO:   04:36:04 - uvicorn.access - 172.17.0.1:54772 - "GET /health HTTP/1.1" 200
INFO:   04:36:04 - docling_serve.app - Health check requested
INFO:   04:36:04 - uvicorn.access - 172.17.0.1:54782 - "GET /health HTTP/1.1" 200
INFO:   04:36:04 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:36:04 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:36:04 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:36:04 - docling_serve.app - [TENANT_ID] Task 3e38ca39-d462-4e95-a1f1-2c1a7948b32e created with tenant_id='default'
INFO:   04:36:04 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 3e38ca39-d462-4e95-a1f1-2c1a7948b32e
INFO:   04:36:04 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:36:04 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 5cd3e3811d2e4d7d996fcfcbbe327b1d
INFO:   04:36:04 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:36:04 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:36:04,346 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:04,346 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:36:04,407 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:04,407 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:36:04,502 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:04,502 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:36:04 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:36:04 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:36:04 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▉    | 454/770 [00:00<00:00, 4535.80it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4878.87it/s]
INFO:   04:36:05 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:36:05 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:36:05 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:36:05 - docling.document_converter - Going to convert document batch...
INFO:   04:36:05 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 5cd3e3811d2e4d7d996fcfcbbe327b1d
INFO:   04:36:05 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:36:06 - docling.document_converter - Finished converting document file in 0.50 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:635: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
WARNING:    04:36:06 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:36:06 - docling_jobkit.convert.results - Processed 1 docs in 0.61 seconds.
INFO:   04:36:06 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 3e38ca39-d462-4e95-a1f1-2c1a7948b32e in 0.61 seconds
INFO:   04:36:08 - uvicorn.access - 172.17.0.1:54782 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:36:08 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:36:08 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:36:08 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:36:08 - docling_serve.app - [TENANT_ID] Task d2cd02b3-5b25-4952-83bc-e3290326e161 created with tenant_id='default'
INFO:   04:36:08 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task d2cd02b3-5b25-4952-83bc-e3290326e161
INFO:   04:36:08 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:36:08 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 5cd3e3811d2e4d7d996fcfcbbe327b1d
INFO:   04:36:08 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:36:08 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:36:08,342 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:08,342 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:36:08,398 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:08,399 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:36:08,505 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:08,506 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:36:08 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:36:08 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:36:08 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▉    | 458/770 [00:00<00:00, 4562.51it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4469.86it/s]
INFO:   04:36:09 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:36:09 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:36:09 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:36:10 - docling.document_converter - Going to convert document batch...
INFO:   04:36:10 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 5cd3e3811d2e4d7d996fcfcbbe327b1d
INFO:   04:36:10 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:36:10 - docling.document_converter - Finished converting document file in 0.98 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:635: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:36:10 - docling_jobkit.convert.results - Processed 1 docs in 0.98 seconds.
INFO:   04:36:10 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job d2cd02b3-5b25-4952-83bc-e3290326e161 in 0.98 seconds
INFO:   04:36:12 - uvicorn.access - 172.17.0.1:54782 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:36:12 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:36:12 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:36:12 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:36:12 - docling_serve.app - [TENANT_ID] Task 863fb628-b6e3-43c3-9022-5c0c85486cbc created with tenant_id='default'
INFO:   04:36:12 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 863fb628-b6e3-43c3-9022-5c0c85486cbc
INFO:   04:36:12 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:36:12 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 9df8df3fc03c33cf6e8bd583e67dce64
INFO:   04:36:12 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:36:12 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:36:12,665 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:12,666 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:36:12,743 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:12,743 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:36:12,817 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:36:12,817 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:36:12 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:36:12 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:36:12 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 481/770 [00:00<00:00, 4799.52it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5078.23it/s]
INFO:   04:36:13 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:36:13 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:36:14 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:36:14 - docling.document_converter - Going to convert document batch...
INFO:   04:36:14 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 9df8df3fc03c33cf6e8bd583e67dce64
INFO:   04:36:14 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:36:14 - docling.document_converter - Finished converting document file in 0.46 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:635: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:36:14 - docling_jobkit.convert.results - Processed 1 docs in 0.52 seconds.
INFO:   04:36:14 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 863fb628-b6e3-43c3-9022-5c0c85486cbc in 0.52 seconds
INFO:   04:36:16 - uvicorn.access - 172.17.0.1:54782 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:36:16 - uvicorn.access - 172.17.0.1:54782 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:36:16 - uvicorn.access - 172.17.0.1:54782 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.29.0#

Click to expand

###### Message

Click to collapse

Tag v1.29.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:34:36 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:34:36 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:34:36 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:34:37 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:34:37 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:34:37 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:34:37 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:34:37 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/health$,/metrics$,/healthz$,/ready$,/livez$,/readyz$)
INFO:   04:34:37 - uvicorn.error - Started server process [1]
INFO:   04:34:37 - uvicorn.error - Waiting for application startup.
INFO:   04:34:43 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:34:44 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:34:44 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:34:44 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:34:44 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:34:44 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash caff667a02b7f037838d32d34acfe0b0
INFO:   04:34:45 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:34:45 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:34:45 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-08-24 04:34:45.062942663 [W:onnxruntime:Default, device_discovery.cc:134 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:34:45 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:34:45,848 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:45,851 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:34:45,995 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:45,995 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:34:46,058 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:46,059 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:34:46 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:34:46 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  11%|█         | 82/770 [00:00<00:02, 305.08it/s]
Loading weights:  15%|█▍        | 113/770 [00:00<00:02, 296.91it/s]
Loading weights:  19%|█▊        | 143/770 [00:00<00:04, 153.11it/s]
Loading weights:  27%|██▋       | 206/770 [00:00<00:02, 244.79it/s]
Loading weights:  32%|███▏      | 243/770 [00:01<00:02, 243.28it/s]
Loading weights:  36%|███▌      | 276/770 [00:01<00:04, 114.45it/s]
Loading weights:  46%|████▋     | 358/770 [00:01<00:02, 197.74it/s]
Loading weights:  78%|███████▊  | 604/770 [00:01<00:00, 525.15it/s]
Loading weights: 100%|██████████| 770/770 [00:01<00:00, 390.40it/s]
INFO:   04:34:48 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:34:49 - uvicorn.error - Application startup complete.
INFO:   04:34:49 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:34:50 - docling_serve.app - Health check requested
INFO:   04:34:50 - uvicorn.access - 172.17.0.1:42490 - "GET /health HTTP/1.1" 200
INFO:   04:34:50 - docling_serve.app - Health check requested
INFO:   04:34:50 - uvicorn.access - 172.17.0.1:42504 - "GET /health HTTP/1.1" 200
INFO:   04:34:50 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:34:50 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:34:50 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:34:50 - docling_serve.app - [TENANT_ID] Task 682b9685-806c-45c8-a9fe-d25e8e98d4ab created with tenant_id='default'
INFO:   04:34:50 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 682b9685-806c-45c8-a9fe-d25e8e98d4ab
INFO:   04:34:50 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:34:51 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash ab12ef348c8fc273bf2be7a49773b303
INFO:   04:34:51 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:34:51 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:34:51,174 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:51,174 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:34:51,241 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:51,241 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:34:51,352 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:51,352 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:34:51 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:34:51 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  68%|██████▊   | 527/770 [00:00<00:00, 5260.46it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4920.17it/s]
INFO:   04:34:52 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:34:53 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:34:53 - docling.document_converter - Going to convert document batch...
INFO:   04:34:53 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:34:53 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:34:53 - docling.document_converter - Finished converting document file in 0.59 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:632: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
WARNING:    04:34:53 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:34:53 - docling_jobkit.convert.results - Processed 1 docs in 0.70 seconds.
INFO:   04:34:53 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 682b9685-806c-45c8-a9fe-d25e8e98d4ab in 0.70 seconds
INFO:   04:34:54 - uvicorn.access - 172.17.0.1:42504 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:34:54 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:34:54 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:34:54 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:34:54 - docling_serve.app - [TENANT_ID] Task 7d8e5113-45ac-4cf5-b285-0d4cdf102889 created with tenant_id='default'
INFO:   04:34:54 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 7d8e5113-45ac-4cf5-b285-0d4cdf102889
INFO:   04:34:54 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:34:55 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash ab12ef348c8fc273bf2be7a49773b303
INFO:   04:34:55 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:34:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:34:55,179 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:55,179 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:34:55,242 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:55,243 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:34:55,322 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:55,323 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:34:55 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:34:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  74%|███████▎  | 567/770 [00:00<00:00, 5668.26it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 7275.40it/s]
INFO:   04:34:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:34:56 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:34:56 - docling.document_converter - Going to convert document batch...
INFO:   04:34:56 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:34:56 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:34:56 - docling.document_converter - Finished converting document file in 0.48 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:632: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:34:56 - docling_jobkit.convert.results - Processed 1 docs in 0.48 seconds.
INFO:   04:34:56 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 7d8e5113-45ac-4cf5-b285-0d4cdf102889 in 0.48 seconds
INFO:   04:34:56 - uvicorn.access - 172.17.0.1:42504 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:34:56 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:34:56 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:34:56 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:34:56 - docling_serve.app - [TENANT_ID] Task a5e72965-82d7-4342-b438-61c21957d2f9 created with tenant_id='default'
INFO:   04:34:56 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task a5e72965-82d7-4342-b438-61c21957d2f9
INFO:   04:34:56 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:34:57 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash b3849eabdc7ecae5f3d54428a0162342
INFO:   04:34:57 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:34:57 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:34:57,220 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:57,221 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:34:57,312 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:57,313 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:34:57,380 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:34:57,380 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:34:57 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:34:57 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  60%|██████    | 465/770 [00:00<00:00, 4647.10it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4664.01it/s]
INFO:   04:34:58 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:34:58 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:34:59 - docling.document_converter - Going to convert document batch...
INFO:   04:34:59 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:34:59 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:34:59 - docling.document_converter - Finished converting document file in 0.46 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:632: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:34:59 - docling_jobkit.convert.results - Processed 1 docs in 0.51 seconds.
INFO:   04:34:59 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job a5e72965-82d7-4342-b438-61c21957d2f9 in 0.51 seconds
INFO:   04:35:00 - uvicorn.access - 172.17.0.1:42504 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:35:01 - uvicorn.access - 172.17.0.1:42504 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:35:01 - uvicorn.access - 172.17.0.1:42504 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.28.0#

Click to expand

###### Message

Click to collapse

Tag v1.28.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:33:02 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:33:02 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:33:02 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:33:03 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:33:03 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:33:03 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:33:03 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/ready$,/metrics$,/healthz$,/health$,/readyz$,/livez$)
INFO:   04:33:03 - uvicorn.error - Started server process [1]
INFO:   04:33:03 - uvicorn.error - Waiting for application startup.
INFO:   04:33:06 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:33:06 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:33:06 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:33:06 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:33:06 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:33:06 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 118619a1d3ed3b201b814dac997d6742
INFO:   04:33:06 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:33:06 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:33:06 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-08-24 04:33:06.934194534 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:33:07 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:33:07,193 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:07,194 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:33:07,253 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:07,254 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:33:07,303 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:07,303 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:33:07 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:33:07 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  64%|██████▍   | 496/770 [00:00<00:00, 4956.62it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5062.16it/s]
INFO:   04:33:07 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:33:08 - uvicorn.error - Application startup complete.
INFO:   04:33:08 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:33:09 - docling_serve.app - Health check requested
INFO:   04:33:09 - uvicorn.access - 172.17.0.1:34402 - "GET /health HTTP/1.1" 200
INFO:   04:33:09 - docling_serve.app - Health check requested
INFO:   04:33:09 - uvicorn.access - 172.17.0.1:34408 - "GET /health HTTP/1.1" 200
INFO:   04:33:09 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:33:09 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:33:09 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:33:09 - docling_serve.app - [TENANT_ID] Task f217b197-b102-45cf-9777-fdcf5ee5a837 created with tenant_id='default'
INFO:   04:33:09 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task f217b197-b102-45cf-9777-fdcf5ee5a837
INFO:   04:33:09 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:33:09 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 83f6e7119a1499def48d21cd30b03869
INFO:   04:33:09 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:33:09 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:33:09,268 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:09,269 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:33:09,329 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:09,329 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:33:09,464 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:09,465 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:33:09 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:33:09 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  63%|██████▎   | 482/770 [00:00<00:00, 4796.78it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4636.77it/s]
INFO:   04:33:10 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:33:11 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:33:11 - docling.document_converter - Going to convert document batch...
INFO:   04:33:11 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:33:11 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:33:11 - docling.document_converter - Finished converting document file in 0.50 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:611: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
WARNING:    04:33:11 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:33:11 - docling_jobkit.convert.results - Processed 1 docs in 0.61 seconds.
INFO:   04:33:11 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job f217b197-b102-45cf-9777-fdcf5ee5a837 in 0.61 seconds
INFO:   04:33:13 - uvicorn.access - 172.17.0.1:34408 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:33:13 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:33:13 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:33:13 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:33:13 - docling_serve.app - [TENANT_ID] Task 3dde82c9-7e60-47bc-a292-6cc0fdd3ce1f created with tenant_id='default'
INFO:   04:33:13 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 3dde82c9-7e60-47bc-a292-6cc0fdd3ce1f
INFO:   04:33:13 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:33:13 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 83f6e7119a1499def48d21cd30b03869
INFO:   04:33:13 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:33:13 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:33:13,255 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:13,255 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:33:13,334 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:13,335 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:33:13,384 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:13,385 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:33:13 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:33:13 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  66%|██████▌   | 510/770 [00:00<00:00, 5077.39it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5633.55it/s]
INFO:   04:33:14 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:33:14 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:33:14 - docling.document_converter - Going to convert document batch...
INFO:   04:33:14 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:33:14 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:33:14 - docling.document_converter - Finished converting document file in 0.44 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:611: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:33:14 - docling_jobkit.convert.results - Processed 1 docs in 0.44 seconds.
INFO:   04:33:14 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 3dde82c9-7e60-47bc-a292-6cc0fdd3ce1f in 0.44 seconds
INFO:   04:33:15 - uvicorn.access - 172.17.0.1:34408 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:33:15 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:33:15 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:33:15 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:33:15 - docling_serve.app - [TENANT_ID] Task 39b932cc-82fe-4766-960f-c7e3560cd00f created with tenant_id='default'
INFO:   04:33:15 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 39b932cc-82fe-4766-960f-c7e3560cd00f
INFO:   04:33:15 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:33:15 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 528ff53b5cc66cd5f7f61ef2cd72a8fc
INFO:   04:33:15 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:33:15 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:33:15,326 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:15,326 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:33:15,399 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:15,400 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:33:15,469 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:33:15,469 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:33:15 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:33:15 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  61%|██████    | 469/770 [00:00<00:00, 4664.08it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4437.54it/s]
INFO:   04:33:16 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:33:16 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:33:16 - docling.document_converter - Going to convert document batch...
INFO:   04:33:16 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:33:16 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:33:16 - docling.document_converter - Finished converting document file in 0.44 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:611: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:33:16 - docling_jobkit.convert.results - Processed 1 docs in 0.49 seconds.
INFO:   04:33:16 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 39b932cc-82fe-4766-960f-c7e3560cd00f in 0.49 seconds
INFO:   04:33:17 - uvicorn.access - 172.17.0.1:34408 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:33:17 - uvicorn.access - 172.17.0.1:34408 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:33:17 - uvicorn.access - 172.17.0.1:34408 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.27.0#

Click to expand

###### Message

Click to collapse

Tag v1.27.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:31:43 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:31:43 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:31:43 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:31:43 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:31:43 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/health$,/readyz$,/ready$,/healthz$,/livez$)
INFO:   04:31:43 - uvicorn.error - Started server process [1]
INFO:   04:31:43 - uvicorn.error - Waiting for application startup.
INFO:   04:31:51 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:31:51 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:31:51 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:31:51 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:31:51 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:31:51 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 118619a1d3ed3b201b814dac997d6742
INFO:   04:31:51 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:31:51 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:31:51 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-08-24 04:31:51.658302610 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:31:52 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:31:52,075 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:52,077 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:31:52,202 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:52,203 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:31:52,323 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:52,324 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:31:52 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:31:52 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  61%|██████    | 467/770 [00:00<00:00, 4654.56it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5143.82it/s]
INFO:   04:31:53 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:31:53 - uvicorn.error - Application startup complete.
INFO:   04:31:53 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:31:54 - docling_serve.app - Health check requested
INFO:   04:31:54 - uvicorn.access - 172.17.0.1:40358 - "GET /health HTTP/1.1" 200
INFO:   04:31:54 - docling_serve.app - Health check requested
INFO:   04:31:54 - uvicorn.access - 172.17.0.1:40360 - "GET /health HTTP/1.1" 200
INFO:   04:31:54 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:31:54 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:31:54 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:31:54 - docling_serve.app - [TENANT_ID] Task 312139e4-666e-4776-bcae-f937f46b4874 created with tenant_id='default'
INFO:   04:31:54 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 312139e4-666e-4776-bcae-f937f46b4874
INFO:   04:31:54 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:31:54 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 83f6e7119a1499def48d21cd30b03869
INFO:   04:31:54 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:31:54 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:31:54,654 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:54,654 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:31:54,731 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:54,733 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:31:54,802 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:54,802 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:31:55 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:31:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 20733.22it/s]
INFO:   04:31:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:31:56 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:31:56 - docling.document_converter - Going to convert document batch...
INFO:   04:31:56 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:31:56 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:31:56 - docling.document_converter - Finished converting document file in 0.55 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:588: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
WARNING:    04:31:56 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:31:56 - docling_jobkit.convert.results - Processed 1 docs in 0.67 seconds.
INFO:   04:31:56 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 312139e4-666e-4776-bcae-f937f46b4874 in 0.67 seconds
INFO:   04:31:58 - uvicorn.access - 172.17.0.1:40360 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:31:58 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:31:58 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:31:58 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:31:58 - docling_serve.app - [TENANT_ID] Task 9ad18297-4c95-491c-8713-0daf142a8fad created with tenant_id='default'
INFO:   04:31:58 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 9ad18297-4c95-491c-8713-0daf142a8fad
INFO:   04:31:58 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:31:58 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 83f6e7119a1499def48d21cd30b03869
INFO:   04:31:58 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:31:58 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:31:58,692 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:58,692 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:31:58,822 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:58,823 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:31:58,901 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:31:58,902 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:31:59 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:31:59 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▉    | 455/770 [00:00<00:00, 4509.71it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4952.72it/s]
INFO:   04:32:00 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:01 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:32:01 - docling.document_converter - Going to convert document batch...
INFO:   04:32:01 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:32:01 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:32:01 - docling.document_converter - Finished converting document file in 0.50 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:588: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:32:01 - docling_jobkit.convert.results - Processed 1 docs in 0.50 seconds.
INFO:   04:32:01 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 9ad18297-4c95-491c-8713-0daf142a8fad in 0.50 seconds
INFO:   04:32:02 - uvicorn.access - 172.17.0.1:40360 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:32:02 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:32:02 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:32:02 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:32:02 - docling_serve.app - [TENANT_ID] Task 6efb0fb6-71b7-47d6-81fa-e5b3e4ea39c0 created with tenant_id='default'
INFO:   04:32:02 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 6efb0fb6-71b7-47d6-81fa-e5b3e4ea39c0
INFO:   04:32:02 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:02 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 528ff53b5cc66cd5f7f61ef2cd72a8fc
INFO:   04:32:02 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:32:02 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:32:02,693 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:32:02,693 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-08-24 04:32:02,779 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:32:02,779 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:32:02,826 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:32:02,826 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:32:02 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:02 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  66%|██████▌   | 509/770 [00:00<00:00, 5086.10it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5163.90it/s]
INFO:   04:32:03 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:04 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:32:04 - docling.document_converter - Going to convert document batch...
INFO:   04:32:04 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:32:04 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:32:04 - docling.document_converter - Finished converting document file in 0.48 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:588: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:32:04 - docling_jobkit.convert.results - Processed 1 docs in 0.53 seconds.
INFO:   04:32:04 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 6efb0fb6-71b7-47d6-81fa-e5b3e4ea39c0 in 0.53 seconds
INFO:   04:32:06 - uvicorn.access - 172.17.0.1:40360 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:32:06 - uvicorn.access - 172.17.0.1:40360 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:32:06 - uvicorn.access - 172.17.0.1:40360 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.26.0#

Click to expand

###### Message

Click to collapse

Tag v1.26.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:30:15 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:30:15 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:30:15 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:30:15 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:30:15 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/livez$,/readyz$,/health$,/healthz$,/metrics$,/ready$)
INFO:   04:30:15 - uvicorn.error - Started server process [1]
INFO:   04:30:15 - uvicorn.error - Waiting for application startup.
INFO:   04:30:18 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:30:18 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:30:18 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:30:18 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:30:18 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:30:18 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c4d6547d10e26a777a530d63ca01a1a1
INFO:   04:30:18 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:30:18 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:30:18 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-08-24 04:30:18.874705432 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:30:19 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:30:19,383 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:19,385 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:30:19,489 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:19,489 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:30:19,535 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:19,535 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:30:19 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:30:19 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 480/770 [00:00<00:00, 4796.88it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4861.90it/s]
INFO:   04:30:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:30:20 - uvicorn.error - Application startup complete.
INFO:   04:30:20 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:30:20 - docling_serve.app - Health check requested
INFO:   04:30:20 - uvicorn.access - 172.17.0.1:39456 - "GET /health HTTP/1.1" 200
INFO:   04:30:20 - docling_serve.app - Health check requested
INFO:   04:30:20 - uvicorn.access - 172.17.0.1:39470 - "GET /health HTTP/1.1" 200
INFO:   04:30:20 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:30:20 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:30:20 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:30:20 - docling_serve.app - [TENANT_ID] Task 51701e21-8620-47e2-8b6a-849b7c241b01 created with tenant_id='default'
INFO:   04:30:20 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 51701e21-8620-47e2-8b6a-849b7c241b01
INFO:   04:30:20 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:30:20 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 4ef9920c0fc80ed745870d6025fc17ac
INFO:   04:30:20 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:30:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:30:20,657 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:20,657 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:30:20,776 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:20,776 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:30:20,824 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:20,824 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:30:20 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:30:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 513/770 [00:00<00:00, 5114.26it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5120.67it/s]
INFO:   04:30:21 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:30:21 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:30:21 - docling.document_converter - Going to convert document batch...
INFO:   04:30:21 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:30:21 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:30:22 - docling.document_converter - Finished converting document file in 0.39 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:556: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
WARNING:    04:30:22 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:30:22 - docling_jobkit.convert.results - Processed 1 docs in 0.51 seconds.
INFO:   04:30:22 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 51701e21-8620-47e2-8b6a-849b7c241b01 in 0.51 seconds
INFO:   04:30:22 - uvicorn.access - 172.17.0.1:39470 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:30:22 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:30:22 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:30:22 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:30:22 - docling_serve.app - [TENANT_ID] Task 372de451-9a35-4d19-befb-899082bfa691 created with tenant_id='default'
INFO:   04:30:22 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 372de451-9a35-4d19-befb-899082bfa691
INFO:   04:30:22 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:30:22 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 4ef9920c0fc80ed745870d6025fc17ac
INFO:   04:30:22 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:30:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:30:22,724 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:22,724 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:30:22,862 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:22,862 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:30:22,973 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:22,973 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:30:23 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:30:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 477/770 [00:00<00:00, 4767.79it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4628.41it/s]
INFO:   04:30:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:30:24 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:30:24 - docling.document_converter - Going to convert document batch...
INFO:   04:30:24 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:30:24 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:30:24 - docling.document_converter - Finished converting document file in 0.35 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:556: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:30:24 - docling_jobkit.convert.results - Processed 1 docs in 0.35 seconds.
INFO:   04:30:24 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 372de451-9a35-4d19-befb-899082bfa691 in 0.35 seconds
INFO:   04:30:24 - uvicorn.access - 172.17.0.1:39470 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:30:24 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:30:24 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:30:24 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:30:24 - docling_serve.app - [TENANT_ID] Task 332a5650-bf60-435d-bc4c-eb9dda568980 created with tenant_id='default'
INFO:   04:30:24 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 332a5650-bf60-435d-bc4c-eb9dda568980
INFO:   04:30:24 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:30:25 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 63e8839d1103b03d45382915dba33c89
INFO:   04:30:25 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:30:25 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:30:25,086 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:25,086 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:30:25,206 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:25,207 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:30:25,293 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:30:25,293 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:30:25 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:30:25 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  68%|██████▊   | 525/770 [00:00<00:00, 5249.19it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5186.26it/s]
INFO:   04:30:25 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:30:26 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:30:26 - docling.document_converter - Going to convert document batch...
INFO:   04:30:26 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:30:26 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:30:26 - docling.document_converter - Finished converting document file in 0.45 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:556: RuntimeWarning: Mean of empty slice
  np.nanmean(
/opt/app-root/lib64/python3.12/site-packages/numpy/lib/_nanfunctions_impl.py:1573: RuntimeWarning: All-NaN slice encountered
  return _nanquantile_unchecked(
INFO:   04:30:26 - docling_jobkit.convert.results - Processed 1 docs in 0.51 seconds.
INFO:   04:30:26 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 332a5650-bf60-435d-bc4c-eb9dda568980 in 0.51 seconds
INFO:   04:30:28 - uvicorn.access - 172.17.0.1:39470 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:30:28 - uvicorn.access - 172.17.0.1:39470 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:30:28 - uvicorn.access - 172.17.0.1:39470 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.25.0#

Click to expand

###### Message

Click to collapse

Tag v1.25.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:29:02 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:02 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:29:02 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:29:02 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:29:02 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/health$,/healthz$,/livez$,/readyz$,/metrics$,/ready$)
INFO:   04:29:02 - uvicorn.error - Started server process [1]
INFO:   04:29:02 - uvicorn.error - Waiting for application startup.
INFO:   04:29:07 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:07 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:29:07 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:07 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:29:07 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:07 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash f446dcc5a1a7c6aff458e30cab474a9e
INFO:   04:29:07 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:07 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-08-24 04:29:07.906223945 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:29:08 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:29:08,387 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:08,390 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:29:08,588 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:08,589 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:29:08,717 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:08,721 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:29:09 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:09 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  48%|████▊     | 371/770 [00:00<00:00, 3706.88it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4226.22it/s]
INFO:   04:29:09 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:11 - uvicorn.error - Application startup complete.
INFO:   04:29:11 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:29:11 - docling_serve.app - Health check requested
INFO:   04:29:11 - uvicorn.access - 172.17.0.1:37232 - "GET /health HTTP/1.1" 200
INFO:   04:29:11 - docling_serve.app - Health check requested
INFO:   04:29:11 - uvicorn.access - 172.17.0.1:37238 - "GET /health HTTP/1.1" 200
INFO:   04:29:11 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:11 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:11 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:11 - docling_serve.app - [TENANT_ID] Task b6694695-21cd-49fa-af9e-bcde1fb1ed30 created with tenant_id='default'
INFO:   04:29:11 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task b6694695-21cd-49fa-af9e-bcde1fb1ed30
INFO:   04:29:11 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:11 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c639c48e78aacd42b6243acf42ab1c3a
INFO:   04:29:11 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:29:11,961 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:11,961 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:29:12,134 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:12,135 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:29:12,184 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:12,184 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:29:12 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:12 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  60%|██████    | 464/770 [00:00<00:00, 4630.26it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4880.44it/s]
INFO:   04:29:12 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:14 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:29:14 - docling.document_converter - Going to convert document batch...
INFO:   04:29:14 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:29:14 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:14 - docling.document_converter - Finished converting document file in 1.13 sec.
WARNING:    04:29:14 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:29:14 - docling_jobkit.convert.results - Processed 1 docs in 1.22 seconds.
INFO:   04:29:14 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job b6694695-21cd-49fa-af9e-bcde1fb1ed30 in 1.22 seconds
INFO:   04:29:15 - uvicorn.access - 172.17.0.1:37238 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:15 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:15 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:15 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:15 - docling_serve.app - [TENANT_ID] Task fe2ccf2a-a2e9-4392-bf65-134142a5b771 created with tenant_id='default'
INFO:   04:29:15 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task fe2ccf2a-a2e9-4392-bf65-134142a5b771
INFO:   04:29:15 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:16 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c639c48e78aacd42b6243acf42ab1c3a
INFO:   04:29:16 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:29:16,296 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:16,297 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:29:16,431 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:16,431 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:29:16,486 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:16,487 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:29:16 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:16 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  53%|█████▎    | 407/770 [00:00<00:00, 4057.56it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 6247.97it/s]
INFO:   04:29:17 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:17 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:29:17 - docling.document_converter - Going to convert document batch...
INFO:   04:29:17 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:29:17 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:18 - docling.document_converter - Finished converting document file in 0.45 sec.
INFO:   04:29:18 - docling_jobkit.convert.results - Processed 1 docs in 0.45 seconds.
INFO:   04:29:18 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job fe2ccf2a-a2e9-4392-bf65-134142a5b771 in 0.45 seconds
INFO:   04:29:19 - uvicorn.access - 172.17.0.1:37238 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:19 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:19 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:19 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:19 - docling_serve.app - [TENANT_ID] Task e354c106-d030-48ce-a465-c9317c1c1221 created with tenant_id='default'
INFO:   04:29:19 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task e354c106-d030-48ce-a465-c9317c1c1221
INFO:   04:29:19 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:19 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash e50b6140ec7ee9be0d0fea6ba20c9795
INFO:   04:29:19 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:29:19,970 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:19,971 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:29:20,069 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:20,069 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:29:20,119 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:29:20,119 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:29:20 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  60%|█████▉    | 459/770 [00:00<00:00, 4588.25it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4694.13it/s]
INFO:   04:29:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:21 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:29:21 - docling.document_converter - Going to convert document batch...
INFO:   04:29:21 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:29:21 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:21 - docling.document_converter - Finished converting document file in 0.44 sec.
INFO:   04:29:21 - docling_jobkit.convert.results - Processed 1 docs in 0.50 seconds.
INFO:   04:29:21 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job e354c106-d030-48ce-a465-c9317c1c1221 in 0.50 seconds
INFO:   04:29:21 - uvicorn.access - 172.17.0.1:37238 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:22 - uvicorn.access - 172.17.0.1:37238 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:29:22 - uvicorn.access - 172.17.0.1:37238 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.24.0#

Click to expand

###### Message

Click to collapse

Tag v1.24.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:27:25 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:27:25 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:27:25 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:27:25 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:27:25 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/ready$,/livez$,/readyz$,/health$,/healthz$)
INFO:   04:27:25 - uvicorn.error - Started server process [1]
INFO:   04:27:25 - uvicorn.error - Waiting for application startup.
INFO:   04:27:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:27:29 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:27:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:27:29 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:27:29 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:27:29 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 1743cd6dac10356c82ba7dfffff4ed02
INFO:   04:27:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:27:29 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-08-24 04:27:29.615945391 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:27:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:27:29,951 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:29,952 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:27:30,055 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:30,055 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:27:30,101 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:30,102 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:27:30 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:27:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▌   | 503/770 [00:00<00:00, 5004.71it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4999.93it/s]
INFO:   04:27:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:27:31 - uvicorn.error - Application startup complete.
INFO:   04:27:31 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:27:31 - docling_serve.app - Health check requested
INFO:   04:27:31 - uvicorn.access - 172.17.0.1:59704 - "GET /health HTTP/1.1" 200
INFO:   04:27:31 - docling_serve.app - Health check requested
INFO:   04:27:31 - uvicorn.access - 172.17.0.1:59710 - "GET /health HTTP/1.1" 200
INFO:   04:27:31 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:27:31 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:27:31 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:27:31 - docling_serve.app - [TENANT_ID] Task 65094f32-cc03-43fc-b25b-3c4520d5e083 created with tenant_id='default'
INFO:   04:27:31 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 65094f32-cc03-43fc-b25b-3c4520d5e083
INFO:   04:27:31 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:27:31 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 447bb8cb2d04f05cb199adb5a03148da
INFO:   04:27:31 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:27:31,414 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:31,414 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:27:31,508 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:31,509 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:27:31,556 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:31,557 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:27:31 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:27:31 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  66%|██████▌   | 508/770 [00:00<00:00, 5037.94it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5023.87it/s]
INFO:   04:27:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:27:32 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:27:32 - docling.document_converter - Going to convert document batch...
INFO:   04:27:32 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:27:32 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:27:33 - docling.document_converter - Finished converting document file in 0.74 sec.
WARNING:    04:27:33 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:27:33 - docling_jobkit.convert.results - Processed 1 docs in 0.82 seconds.
INFO:   04:27:33 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 65094f32-cc03-43fc-b25b-3c4520d5e083 in 0.82 seconds
INFO:   04:27:33 - uvicorn.access - 172.17.0.1:59710 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:27:33 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:27:33 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:27:33 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:27:33 - docling_serve.app - [TENANT_ID] Task 13f994da-71df-4ffa-8e63-08834f819399 created with tenant_id='default'
INFO:   04:27:33 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 13f994da-71df-4ffa-8e63-08834f819399
INFO:   04:27:33 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:27:33 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 447bb8cb2d04f05cb199adb5a03148da
INFO:   04:27:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:27:33,454 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:33,454 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:27:33,556 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:33,556 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:27:33,603 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:33,603 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:27:33 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:27:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  63%|██████▎   | 488/770 [00:00<00:00, 4871.76it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4917.29it/s]
INFO:   04:27:34 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:27:34 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:27:34 - docling.document_converter - Going to convert document batch...
INFO:   04:27:34 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:27:34 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:27:34 - docling.document_converter - Finished converting document file in 0.30 sec.
INFO:   04:27:34 - docling_jobkit.convert.results - Processed 1 docs in 0.30 seconds.
INFO:   04:27:34 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 13f994da-71df-4ffa-8e63-08834f819399 in 0.30 seconds
INFO:   04:27:35 - uvicorn.access - 172.17.0.1:59710 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:27:35 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:27:35 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:27:35 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:27:35 - docling_serve.app - [TENANT_ID] Task 202cc3db-78e5-4566-863b-e4f75c2c9bfe created with tenant_id='default'
INFO:   04:27:35 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 202cc3db-78e5-4566-863b-e4f75c2c9bfe
INFO:   04:27:35 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:27:35 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 6102b5a2a09ce02f29b31fc27cc3d04f
INFO:   04:27:35 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:27:35,457 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:35,457 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:27:35,565 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:35,565 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:27:35,628 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:27:35,629 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:27:35 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:27:35 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  64%|██████▍   | 494/770 [00:00<00:00, 4925.82it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4972.41it/s]
INFO:   04:27:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:27:36 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:27:36 - docling.document_converter - Going to convert document batch...
INFO:   04:27:36 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:27:36 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:27:36 - docling.document_converter - Finished converting document file in 0.29 sec.
INFO:   04:27:36 - docling_jobkit.convert.results - Processed 1 docs in 0.34 seconds.
INFO:   04:27:36 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 202cc3db-78e5-4566-863b-e4f75c2c9bfe in 0.34 seconds
INFO:   04:27:37 - uvicorn.access - 172.17.0.1:59710 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:27:37 - uvicorn.access - 172.17.0.1:59710 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:27:37 - uvicorn.access - 172.17.0.1:59710 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.23.0#

Click to expand

###### Message

Click to collapse

Tag v1.23.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:26:21 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:21 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:26:21 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:26:21 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:26:21 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/health$,/ready$,/livez$,/healthz$,/metrics$,/readyz$)
INFO:   04:26:22 - uvicorn.error - Started server process [1]
INFO:   04:26:22 - uvicorn.error - Waiting for application startup.
INFO:   04:26:28 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:28 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:26:28 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:28 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:26:28 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:28 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 1743cd6dac10356c82ba7dfffff4ed02
INFO:   04:26:28 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:28 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-08-24 04:26:28.781986984 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:26:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:26:29,235 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:29,237 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:26:29,401 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:29,401 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:26:29,468 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:29,468 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:26:29 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 513/770 [00:00<00:00, 5125.90it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5071.51it/s]
INFO:   04:26:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:31 - uvicorn.error - Application startup complete.
INFO:   04:26:31 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:26:32 - docling_serve.app - Health check requested
INFO:   04:26:32 - uvicorn.access - 172.17.0.1:38486 - "GET /health HTTP/1.1" 200
INFO:   04:26:32 - docling_serve.app - Health check requested
INFO:   04:26:32 - uvicorn.access - 172.17.0.1:38494 - "GET /health HTTP/1.1" 200
INFO:   04:26:32 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:26:32 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:26:32 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:26:32 - docling_serve.app - [TENANT_ID] Task 5adecfbb-0ba8-4448-a8e8-8c46329960ef created with tenant_id='default'
INFO:   04:26:32 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 5adecfbb-0ba8-4448-a8e8-8c46329960ef
INFO:   04:26:32 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:32 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 447bb8cb2d04f05cb199adb5a03148da
INFO:   04:26:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:26:32,272 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:32,272 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:26:32,394 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:32,394 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:26:32,462 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:32,462 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:26:32 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  63%|██████▎   | 484/770 [00:00<00:00, 4832.10it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4732.19it/s]
INFO:   04:26:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:33 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:33 - docling.document_converter - Going to convert document batch...
INFO:   04:26:33 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:26:33 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:34 - docling.document_converter - Finished converting document file in 0.53 sec.
WARNING:    04:26:34 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:26:34 - docling_jobkit.convert.results - Processed 1 docs in 0.64 seconds.
INFO:   04:26:34 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 5adecfbb-0ba8-4448-a8e8-8c46329960ef in 0.64 seconds
INFO:   04:26:36 - uvicorn.access - 172.17.0.1:38494 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:26:36 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:26:36 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:26:36 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:26:36 - docling_serve.app - [TENANT_ID] Task f68a8350-673d-44a0-9f2b-4108fdcd65d9 created with tenant_id='default'
INFO:   04:26:36 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task f68a8350-673d-44a0-9f2b-4108fdcd65d9
INFO:   04:26:36 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:36 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 447bb8cb2d04f05cb199adb5a03148da
INFO:   04:26:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:26:36,304 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:36,304 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:26:36,458 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:36,459 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:26:36,516 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:36,516 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:26:36 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  61%|██████    | 467/770 [00:00<00:00, 4631.83it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4848.19it/s]
INFO:   04:26:37 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:38 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:38 - docling.document_converter - Going to convert document batch...
INFO:   04:26:38 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:26:38 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:38 - docling.document_converter - Finished converting document file in 0.49 sec.
INFO:   04:26:38 - docling_jobkit.convert.results - Processed 1 docs in 0.49 seconds.
INFO:   04:26:38 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job f68a8350-673d-44a0-9f2b-4108fdcd65d9 in 0.49 seconds
INFO:   04:26:40 - uvicorn.access - 172.17.0.1:38494 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:26:40 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:26:40 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:26:40 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:26:40 - docling_serve.app - [TENANT_ID] Task b25fb668-551b-455a-82e0-1719d35da987 created with tenant_id='default'
INFO:   04:26:40 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task b25fb668-551b-455a-82e0-1719d35da987
INFO:   04:26:40 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:40 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 6102b5a2a09ce02f29b31fc27cc3d04f
INFO:   04:26:40 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:26:40,310 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:40,310 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:26:40,456 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:40,456 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:26:40,507 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:26:40,508 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:26:40 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:40 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  55%|█████▌    | 425/770 [00:00<00:00, 4249.10it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4178.18it/s]
INFO:   04:26:41 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:41 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:42 - docling.document_converter - Going to convert document batch...
INFO:   04:26:42 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:26:42 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:42 - docling.document_converter - Finished converting document file in 0.49 sec.
INFO:   04:26:42 - docling_jobkit.convert.results - Processed 1 docs in 0.53 seconds.
INFO:   04:26:42 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job b25fb668-551b-455a-82e0-1719d35da987 in 0.53 seconds
INFO:   04:26:42 - uvicorn.access - 172.17.0.1:38494 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:26:42 - uvicorn.access - 172.17.0.1:38494 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:26:42 - uvicorn.access - 172.17.0.1:38494 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.22.1#

Click to expand

###### Message

Click to collapse

Tag v1.22.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:24:43 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:24:43 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:24:43 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:24:43 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:24:43 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/ready$,/livez$,/readyz$,/healthz$,/health$)
INFO:   04:24:43 - uvicorn.error - Started server process [1]
INFO:   04:24:43 - uvicorn.error - Waiting for application startup.
INFO:   04:24:46 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:24:46 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:24:46 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:24:46 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:24:46 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:24:46 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 1743cd6dac10356c82ba7dfffff4ed02
INFO:   04:24:46 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:24:46 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-08-24 04:24:47.016849284 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:24:47 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:24:47,307 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:47,308 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:24:47,399 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:47,399 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:24:47,444 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:47,444 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:24:47 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:24:47 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▌   | 502/770 [00:00<00:00, 4997.89it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4894.21it/s]
INFO:   04:24:48 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:24:48 - uvicorn.error - Application startup complete.
INFO:   04:24:48 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:24:48 - docling_serve.app - Health check requested
INFO:   04:24:48 - uvicorn.access - 172.17.0.1:33290 - "GET /health HTTP/1.1" 200
INFO:   04:24:48 - docling_serve.app - Health check requested
INFO:   04:24:48 - uvicorn.access - 172.17.0.1:33300 - "GET /health HTTP/1.1" 200
INFO:   04:24:48 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:24:48 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:24:48 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:24:48 - docling_serve.app - [TENANT_ID] Task 6593b3c8-360f-430e-b869-4b088e0871a1 created with tenant_id='default'
INFO:   04:24:48 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 6593b3c8-360f-430e-b869-4b088e0871a1
INFO:   04:24:48 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:24:48 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a1cb01c48360ff351eb870dc682aac59
INFO:   04:24:48 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:24:48,763 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:48,763 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:24:48,863 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:48,863 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:24:48,931 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:48,932 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:24:49 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:24:49 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  69%|██████▊   | 528/770 [00:00<00:00, 5268.51it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5096.87it/s]
INFO:   04:24:49 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:24:50 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:24:50 - docling.document_converter - Going to convert document batch...
INFO:   04:24:50 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:24:50 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:24:50 - docling.document_converter - Finished converting document file in 0.43 sec.
WARNING:    04:24:50 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:24:50 - docling_jobkit.convert.results - Processed 1 docs in 0.78 seconds.
INFO:   04:24:50 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 6593b3c8-360f-430e-b869-4b088e0871a1 in 0.78 seconds
INFO:   04:24:50 - uvicorn.access - 172.17.0.1:33300 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:24:50 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:24:50 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:24:50 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:24:50 - docling_serve.app - [TENANT_ID] Task 1e79ab45-503d-40a4-b24e-edf70eb551e5 created with tenant_id='default'
INFO:   04:24:50 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 1e79ab45-503d-40a4-b24e-edf70eb551e5
INFO:   04:24:50 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:24:50 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a1cb01c48360ff351eb870dc682aac59
INFO:   04:24:50 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:24:50,769 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:50,769 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:24:50,863 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:50,863 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:24:50,912 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:50,912 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:24:51 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:24:51 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 514/770 [00:00<00:00, 5105.18it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5161.75it/s]
INFO:   04:24:51 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:24:51 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:24:51 - docling.document_converter - Going to convert document batch...
INFO:   04:24:51 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:24:51 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:24:52 - docling.document_converter - Finished converting document file in 0.30 sec.
INFO:   04:24:52 - docling_jobkit.convert.results - Processed 1 docs in 0.31 seconds.
INFO:   04:24:52 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 1e79ab45-503d-40a4-b24e-edf70eb551e5 in 0.31 seconds
INFO:   04:24:52 - uvicorn.access - 172.17.0.1:33300 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:24:52 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:24:52 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:24:52 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:24:52 - docling_serve.app - [TENANT_ID] Task 2a63cf46-2ebf-4503-8a37-e9720986d48a created with tenant_id='default'
INFO:   04:24:52 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 2a63cf46-2ebf-4503-8a37-e9720986d48a
INFO:   04:24:52 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:24:52 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash fb1751b09de6bdbe35ede48676ca8f82
INFO:   04:24:52 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:24:52,779 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:52,779 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:24:52,879 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:52,879 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:24:52,925 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:52,925 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:24:53 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:24:53 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  66%|██████▌   | 508/770 [00:00<00:00, 5078.01it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5017.80it/s]
INFO:   04:24:53 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:24:53 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:24:53 - docling.document_converter - Going to convert document batch...
INFO:   04:24:53 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:24:53 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:24:54 - docling.document_converter - Finished converting document file in 0.29 sec.
INFO:   04:24:54 - docling_jobkit.convert.results - Processed 1 docs in 0.36 seconds.
INFO:   04:24:54 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 2a63cf46-2ebf-4503-8a37-e9720986d48a in 0.36 seconds
INFO:   04:24:54 - uvicorn.access - 172.17.0.1:33300 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:24:55 - uvicorn.access - 172.17.0.1:33300 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:24:55 - uvicorn.access - 172.17.0.1:33300 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.22.0#

Click to expand

###### Message

Click to collapse

Tag v1.22.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:   04:23:47 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:47 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:23:47 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:23:47 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:23:47 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/readyz$,/health$,/metrics$,/healthz$,/ready$,/livez$)
INFO:   04:23:47 - uvicorn.error - Started server process [1]
INFO:   04:23:47 - uvicorn.error - Waiting for application startup.
INFO:   04:23:54 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:54 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:23:54 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:54 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:23:54 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:54 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash bcb2e23b24bd94a42b5c00d662a7c1cf
INFO:   04:23:54 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:54 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-08-24 04:23:54.371085171 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
INFO:   04:23:54 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:23:54,808 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:23:54,810 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:23:54,971 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:23:54,971 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:23:55,044 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:23:55,044 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:23:55 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  13%|█▎        | 102/770 [00:00<00:00, 744.88it/s]
Loading weights:  23%|██▎       | 177/770 [00:00<00:03, 184.16it/s]
Loading weights:  28%|██▊       | 215/770 [00:00<00:02, 212.28it/s]
Loading weights:  33%|███▎      | 252/770 [00:01<00:02, 222.03it/s]
Loading weights:  67%|██████▋   | 515/770 [00:01<00:00, 679.00it/s]
Loading weights: 100%|██████████| 770/770 [00:01<00:00, 607.67it/s]
INFO:   04:23:57 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:58 - uvicorn.error - Application startup complete.
INFO:   04:23:58 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:23:58 - docling_serve.app - Health check requested
INFO:   04:23:58 - uvicorn.access - 172.17.0.1:49690 - "GET /health HTTP/1.1" 200
INFO:   04:23:58 - docling_serve.app - Health check requested
INFO:   04:23:58 - uvicorn.access - 172.17.0.1:49692 - "GET /health HTTP/1.1" 200
INFO:   04:23:58 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:23:58 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:23:58 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:23:58 - docling_serve.app - [TENANT_ID] Task 65210509-bf9b-416a-9e29-ffd992d21053 created with tenant_id='default'
INFO:   04:23:58 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 65210509-bf9b-416a-9e29-ffd992d21053
INFO:   04:23:58 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:58 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c835e945376007df084ecda847e91f5b
INFO:   04:23:58 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:23:58,364 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:23:58,365 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:23:58,471 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:23:58,471 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:23:58,519 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:23:58,519 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:23:58 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:58 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  58%|█████▊    | 448/770 [00:00<00:00, 4475.42it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4618.35it/s]
INFO:   04:23:59 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:59 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:24:00 - docling.document_converter - Going to convert document batch...
INFO:   04:24:00 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:24:00 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:24:00 - docling.document_converter - Finished converting document file in 0.53 sec.
WARNING:    04:24:00 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:24:00 - docling_jobkit.convert.results - Processed 1 docs in 0.98 seconds.
INFO:   04:24:00 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 65210509-bf9b-416a-9e29-ffd992d21053 in 0.98 seconds
INFO:   04:24:02 - uvicorn.access - 172.17.0.1:49692 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:24:02 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:24:02 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:24:02 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:24:02 - docling_serve.app - [TENANT_ID] Task 59f5c2a6-130e-4d4c-95ab-df3e338db81a created with tenant_id='default'
INFO:   04:24:02 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 59f5c2a6-130e-4d4c-95ab-df3e338db81a
INFO:   04:24:02 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:24:02 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c835e945376007df084ecda847e91f5b
INFO:   04:24:02 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:24:02,565 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:02,565 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:24:02,702 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:02,702 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:24:02,807 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:02,807 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:24:02 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:24:02 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▉    | 455/770 [00:00<00:00, 4544.26it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4589.72it/s]
INFO:   04:24:03 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:24:04 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:24:04 - docling.document_converter - Going to convert document batch...
INFO:   04:24:04 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:24:04 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:24:04 - docling.document_converter - Finished converting document file in 0.48 sec.
INFO:   04:24:04 - docling_jobkit.convert.results - Processed 1 docs in 0.49 seconds.
INFO:   04:24:04 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 59f5c2a6-130e-4d4c-95ab-df3e338db81a in 0.49 seconds
INFO:   04:24:04 - uvicorn.access - 172.17.0.1:49692 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:24:04 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:24:04 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:24:04 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:24:04 - docling_serve.app - [TENANT_ID] Task 937e1466-5965-4f10-a5de-a61bcfefe76c created with tenant_id='default'
INFO:   04:24:04 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 937e1466-5965-4f10-a5de-a61bcfefe76c
INFO:   04:24:04 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:24:04 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 3eabf304acb4a56e9ef52b02ef69bccb
INFO:   04:24:04 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-08-24 04:24:04,566 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:04,566 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:24:04,684 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:04,684 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:24:04,732 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:24:04,732 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:24:04 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:24:04 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 515/770 [00:00<00:00, 5135.94it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4837.62it/s]
INFO:   04:24:05 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:24:06 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:24:06 - docling.document_converter - Going to convert document batch...
INFO:   04:24:06 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:24:06 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:24:06 - docling.document_converter - Finished converting document file in 0.46 sec.
INFO:   04:24:06 - docling_jobkit.convert.results - Processed 1 docs in 0.52 seconds.
INFO:   04:24:06 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 937e1466-5965-4f10-a5de-a61bcfefe76c in 0.52 seconds
INFO:   04:24:06 - uvicorn.access - 172.17.0.1:49692 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:24:06 - uvicorn.access - 172.17.0.1:49692 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:24:06 - uvicorn.access - 172.17.0.1:49692 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.21.0#

Click to expand

###### Message

Click to collapse

Tag v1.21.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
2026-08-24 04:21:52.000896386 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:21:52,295 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:52,296 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:21:52,384 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:52,384 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:21:52,430 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:52,431 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 515/770 [00:00<00:00, 5146.79it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5207.81it/s]
[INFO] 2026-08-24 04:21:54,080 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:54,081 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:21:54,171 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:54,171 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:21:54,242 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:54,242 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  68%|██████▊   | 527/770 [00:00<00:00, 5252.31it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5122.35it/s]
WARNING:    04:21:55 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
WARNING:    04:21:55 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
[INFO] 2026-08-24 04:21:56,373 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:56,373 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:21:56,465 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:56,465 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:21:56,511 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:56,511 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 513/770 [00:00<00:00, 5127.27it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5175.60it/s]
[INFO] 2026-08-24 04:21:58,115 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:58,115 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:21:58,217 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:58,217 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:21:58,263 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:58,263 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  68%|██████▊   | 522/770 [00:00<00:00, 5181.90it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5199.50it/s]

ghcr.io/docling-project/docling-serve:v1.20.0#

Click to expand

###### Message

Click to collapse

Tag v1.20.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:21:06.280961521 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:21:06,695 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:06,698 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:21:06,830 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:06,831 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:21:06,883 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:06,883 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▌   | 504/770 [00:00<00:00, 5024.41it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4860.30it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:36006 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:36012 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:21:08,549 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:08,550 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:21:08,675 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:08,676 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:21:08,726 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:08,726 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  63%|██████▎   | 486/770 [00:00<00:00, 4854.43it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4546.74it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:36012 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:21:10,863 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:10,864 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:21:10,986 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:10,987 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:21:11,038 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:11,039 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  71%|███████▏  | 550/770 [00:00<00:00, 5423.55it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5919.19it/s]
INFO:     172.17.0.1:36012 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:21:14,885 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:14,885 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:21:15,043 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:15,044 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:21:15,117 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:21:15,117 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▌   | 504/770 [00:00<00:00, 5036.14it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4898.08it/s]
INFO:     172.17.0.1:36012 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36012 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:36012 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.19.0#

Click to expand

###### Message

Click to collapse

Tag v1.19.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:19:12.133166337 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:19:12,436 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:12,437 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:19:12,537 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:12,537 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:19:12,592 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:12,592 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 515/770 [00:00<00:00, 5147.86it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5149.73it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:49214 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:49220 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:19:13,847 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:13,847 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:19:13,936 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:13,936 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:19:14,006 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:14,007 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▍   | 499/770 [00:00<00:00, 4970.59it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5017.16it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:49220 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:19:15,858 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:15,858 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:19:15,950 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:15,950 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:19:15,996 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:15,996 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  61%|██████    | 467/770 [00:00<00:00, 4657.63it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4807.27it/s]
INFO:     172.17.0.1:49220 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:19:17,870 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:17,870 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:19:17,967 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:17,967 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:19:18,014 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:19:18,015 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  64%|██████▎   | 489/770 [00:00<00:00, 4886.68it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5004.73it/s]
INFO:     172.17.0.1:49220 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49220 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:49220 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.18.0#

Click to expand

###### Message

Click to collapse

Tag v1.18.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:18:31.324594349 [W:onnxruntime:Default, device_discovery.cc:133 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename "5620e0c7-8062-4dce-aeb7-520c7ef76171" did not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:18:31,782 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:31,786 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:18:32,195 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:32,196 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:18:32,501 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:32,501 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  56%|█████▌    | 428/770 [00:00<00:00, 4278.93it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4429.55it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:51424 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:51428 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:18:34,679 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:34,679 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:18:34,839 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:34,840 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:18:34,915 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:34,915 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 481/770 [00:00<00:00, 4774.82it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5637.85it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:51428 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:18:38,669 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:38,670 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:18:38,756 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:38,756 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:18:38,803 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:38,803 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 518/770 [00:00<00:00, 5120.89it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5177.68it/s]
INFO:     172.17.0.1:51428 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:18:40,677 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:40,677 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:18:40,767 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:40,767 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:18:40,811 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:18:40,812 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▍   | 497/770 [00:00<00:00, 4957.35it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4791.58it/s]
INFO:     172.17.0.1:51428 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:51428 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:51428 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.17.0#

Click to expand

###### Message

Click to collapse

Tag v1.17.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:16:22.972605981 [W:onnxruntime:Default, device_discovery.cc:132 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:16:23,640 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:23,642 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:16:23,754 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:23,754 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:16:23,800 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:23,801 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▉    | 458/770 [00:00<00:00, 4577.87it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4553.99it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:36988 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:36990 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:16:26,190 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:26,190 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:16:26,314 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:26,314 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:16:26,407 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:26,407 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▉    | 456/770 [00:00<00:00, 4547.63it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4970.75it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:36990 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:16:30,203 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:30,203 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:16:30,304 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:30,305 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:16:30,350 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:30,350 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 514/770 [00:00<00:00, 5139.86it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4892.69it/s]
INFO:     172.17.0.1:36990 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:16:32,222 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:32,222 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-08-24 04:16:32,312 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:32,312 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-08-24 04:16:32,359 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:32,360 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▍   | 497/770 [00:00<00:00, 4959.96it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4969.41it/s]
INFO:     172.17.0.1:36990 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36990 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:36990 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.16.1#

Click to expand

###### Message

Click to collapse

Tag v1.16.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:15:55.062622847 [W:onnxruntime:Default, device_discovery.cc:132 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:15:55,834 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:15:55,836 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:15:56,002 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:15:56,002 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:15:56,056 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:15:56,057 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  70%|███████   | 539/770 [00:00<00:00, 5359.63it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 6443.18it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:39744 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39746 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:15:58,823 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:15:58,823 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:15:58,900 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:15:58,901 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:15:58,942 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:15:58,942 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  72%|███████▏  | 554/770 [00:00<00:00, 5526.92it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5473.63it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39746 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:16:02,830 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:02,830 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:16:02,953 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:02,953 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:16:03,033 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:03,034 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  54%|█████▍    | 419/770 [00:00<00:00, 4181.90it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5502.40it/s]
INFO:     172.17.0.1:39746 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-08-24 04:16:06,890 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:06,890 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:16:06,995 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:06,995 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:16:07,065 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:16:07,065 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  63%|██████▎   | 485/770 [00:00<00:00, 4803.14it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4704.36it/s]
INFO:     172.17.0.1:39746 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39746 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39746 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.15.0#

Click to expand

###### Message

Click to collapse

Tag v1.15.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:13:39.092736746 [W:onnxruntime:Default, device_discovery.cc:132 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:13:39,584 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:13:39,585 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:13:39,680 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:13:39,681 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:13:39,736 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:13:39,736 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:49088 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:49104 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:49104 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49104 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49104 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49104 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:49104 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.14.3#

Click to expand

###### Message

Click to collapse

Tag v1.14.3 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:12:22.815527547 [W:onnxruntime:Default, device_discovery.cc:131 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:12:23,570 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:12:23,572 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:12:23,735 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:12:23,735 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:12:23,779 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:12:23,780 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:53236 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:53240 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:53240 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53240 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53240 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53240 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:53240 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.14.2#

Click to expand

###### Message

Click to collapse

Tag v1.14.2 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:10:25.071373725 [W:onnxruntime:Default, device_discovery.cc:131 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:10:25,548 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:10:25,550 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:10:25,639 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:10:25,640 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:10:25,691 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:10:25,692 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:46158 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:46162 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:46162 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:46162 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:46162 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:46162 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:46162 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.14.1#

Click to expand

###### Message

Click to collapse

Tag v1.14.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:09:17.373433657 [W:onnxruntime:Default, device_discovery.cc:131 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:09:18,099 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:09:18,101 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:09:18,252 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:09:18,253 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:09:18,314 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:09:18,314 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:48502 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:48518 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:48518 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48518 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48518 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48518 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:48518 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.14.0#

Click to expand

###### Message

Click to collapse

Tag v1.14.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:07:18.702801166 [W:onnxruntime:Default, device_discovery.cc:131 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:07:19,210 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:07:19,212 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:07:19,300 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:07:19,300 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:07:19,350 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:07:19,350 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:41930 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:41934 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:41934 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:41934 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:41934 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:41934 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:41934 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.13.1#

Click to expand

###### Message

Click to collapse

Tag v1.13.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:06:08.091739704 [W:onnxruntime:Default, device_discovery.cc:131 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:06:08,805 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:06:08,807 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:06:08,998 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:06:08,998 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:06:09,061 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:06:09,062 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:45184 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:45186 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:45186 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45186 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45186 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45186 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:45186 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.13.0#

Click to expand

###### Message

Click to collapse

Tag v1.13.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:04:04.625075014 [W:onnxruntime:Default, device_discovery.cc:131 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:04:05,126 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:04:05,128 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:04:05,238 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:04:05,238 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:04:05,291 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:04:05,291 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:39318 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39330 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39330 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39330 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39330 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39330 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39330 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.12.0#

Click to expand

###### Message

Click to collapse

Tag v1.12.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
2026-08-24 04:02:51.287914222 [W:onnxruntime:Default, device_discovery.cc:131 GetPciBusId] Skipping pci_bus_id for PCI path at "/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0004:00/MSFT1000:00/5620e0c7-8062-4dce-aeb7-520c7ef76171" because filename ""5620e0c7-8062-4dce-aeb7-520c7ef76171"" dit not match expected pattern of [0-9a-f]+:[0-9a-f]+:[0-9a-f]+[.][0-9a-f]+[m
[INFO] 2026-08-24 04:02:52,666 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:02:52,694 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:02:53,052 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:02:53,052 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:02:53,117 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:02:53,117 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:39344 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39360 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39360 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39360 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39360 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39360 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39360 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.11.0#

Click to expand

###### Message

Click to collapse

Tag v1.11.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
[INFO] 2026-08-24 04:00:51,177 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:00:51,179 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 04:00:51,267 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:00:51,267 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 04:00:51,303 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 04:00:51,303 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:36332 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:36342 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:36342 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36342 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36342 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36342 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:36342 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.10.0#

Click to expand

###### Message

Click to collapse

Tag v1.10.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
[INFO] 2026-08-24 03:59:35,799 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:59:35,802 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 03:59:35,941 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:59:35,942 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 03:59:36,022 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:59:36,022 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:51426 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:51436 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:51436 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:51436 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:51436 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:51436 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:51436 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.9.0#

Click to expand

###### Message

Click to collapse

Tag v1.9.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
[INFO] 2026-08-24 03:57:26,536 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:57:26,538 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 03:57:26,668 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:57:26,669 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 03:57:26,728 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:57:26,728 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:38920 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:38934 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:38934 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:38934 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:38934 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:38934 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:38934 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.8.0#

Click to expand

###### Message

Click to collapse

Tag v1.8.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
[INFO] 2026-08-24 03:55:30,449 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:55:30,452 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 03:55:30,580 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:55:30,581 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 03:55:30,674 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:55:30,674 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:37240 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:37244 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:37244 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:37244 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:37244 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:37244 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:37244 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.7.2#

Click to expand

###### Message

Click to collapse

Tag v1.7.2 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
[INFO] 2026-08-24 03:53:21,012 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:53:21,014 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 03:53:21,110 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:53:21,110 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 03:53:21,151 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:53:21,152 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:43728 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:43738 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:43738 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43738 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43738 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43738 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:43738 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.7.1#

Click to expand

###### Message

Click to collapse

Tag v1.7.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
[INFO] 2026-08-24 03:51:37,937 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:51:37,940 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-08-24 03:51:38,058 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:51:38,058 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/cls/ch_ppocr_mobile_v2.0_cls_infer.onnx
[INFO] 2026-08-24 03:51:38,134 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-08-24 03:51:38,134 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_infer.onnx
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:52882 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:52890 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:52890 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:52890 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:52890 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:52890 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:52890 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.7.0#

Click to expand

###### Message

Click to collapse

Tag v1.7.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:41904 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:41910 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:41910 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:41910 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:41910 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:41910 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:41910 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.6.0#

Click to expand

###### Message

Click to collapse

Tag v1.6.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:49094 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:49096 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:49096 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49096 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49096 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49096 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:49096 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.5.1#

Click to expand

###### Message

Click to collapse

Tag v1.5.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:36484 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:36498 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:36498 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36498 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36498 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36498 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:36498 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.5.0#

Click to expand

###### Message

Click to collapse

Tag v1.5.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:45202 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:45214 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:45214 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45214 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45214 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45214 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:45214 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.4.1#

Click to expand

###### Message

Click to collapse

Tag v1.4.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:53514 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:53530 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:53530 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53530 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53530 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53530 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:53530 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.4.0#

Click to expand

###### Message

Click to collapse

Tag v1.4.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:49088 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:49090 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:49090 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49090 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49090 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49090 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:49090 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.3.1#

Click to expand

###### Message

Click to collapse

Tag v1.3.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:48316 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:48328 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:48328 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48328 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48328 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48328 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:48328 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.3.0#

Click to expand

###### Message

Click to collapse

Tag v1.3.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:49796 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:49808 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:49808 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49808 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49808 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:49808 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:49808 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.2.2#

Click to expand

###### Message

Click to collapse

Tag v1.2.2 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:56342 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:56346 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:56346 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:56346 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:56346 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:56346 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:56346 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.2.1#

Click to expand

###### Message

Click to collapse

Tag v1.2.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:39212 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39214 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39214 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39214 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39214 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39214 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39214 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.2.0#

Click to expand

###### Message

Click to collapse

Tag v1.2.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:60390 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:60402 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:60402 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:60402 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:60402 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:60402 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:60402 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.1.0#

Click to expand

###### Message

Click to collapse

Tag v1.1.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:59526 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:59534 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:59534 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:59534 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:59534 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:59534 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:59534 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.0.1#

Click to expand

###### Message

Click to collapse

Tag v1.0.1 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:59824 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:59836 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:59836 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:59836 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:59836 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:59836 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:59836 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

ghcr.io/docling-project/docling-serve:v1.0.0#

Click to expand

###### Message

Click to collapse

Tag v1.0.0 is ok

###### Docling server logs

click to expand

Starting production server 🚀

Server started at http://0.0.0.0:5001
Documentation at http://0.0.0.0:5001/docs
Scalar docs at http://0.0.0.0:5001/scalar

Logs:
INFO:     Started server process [1]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:45686 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:45696 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:45696 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45696 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45696 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:45696 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:45696 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK

Docling Serve API#

When to use this module#

Installation#

Quick start#

Core concepts#

The DoclingServeApi interface#

Extension Points#

Requests: ConvertDocumentRequest#

Responses: InBodyConvertDocumentResponse, PreSignedUrlConvertDocumentResponse, PreSignedUrlConvertResponse, ZipArchiveConvertDocumentResponse, and DocumentResponse#

Presigned URL target example#

Health checks#

Error handling#

Validation errors#

Logging and builders#

Version compatibility#

Results for ghcr.io/docling-project/docling-serve as of 2026-08-24T04:37:21.526667782Z#

Details#

ghcr.io/docling-project/docling-serve:v1.31.0#

ghcr.io/docling-project/docling-serve:v1.30.0#

ghcr.io/docling-project/docling-serve:v1.29.0#

ghcr.io/docling-project/docling-serve:v1.28.0#

ghcr.io/docling-project/docling-serve:v1.27.0#

ghcr.io/docling-project/docling-serve:v1.26.0#

ghcr.io/docling-project/docling-serve:v1.25.0#

ghcr.io/docling-project/docling-serve:v1.24.0#

ghcr.io/docling-project/docling-serve:v1.23.0#

ghcr.io/docling-project/docling-serve:v1.22.1#

ghcr.io/docling-project/docling-serve:v1.22.0#

ghcr.io/docling-project/docling-serve:v1.21.0#

ghcr.io/docling-project/docling-serve:v1.20.0#

ghcr.io/docling-project/docling-serve:v1.19.0#

ghcr.io/docling-project/docling-serve:v1.18.0#

ghcr.io/docling-project/docling-serve:v1.17.0#

ghcr.io/docling-project/docling-serve:v1.16.1#

ghcr.io/docling-project/docling-serve:v1.15.0#

ghcr.io/docling-project/docling-serve:v1.14.3#

ghcr.io/docling-project/docling-serve:v1.14.2#

ghcr.io/docling-project/docling-serve:v1.14.1#

ghcr.io/docling-project/docling-serve:v1.14.0#

ghcr.io/docling-project/docling-serve:v1.13.1#

ghcr.io/docling-project/docling-serve:v1.13.0#

ghcr.io/docling-project/docling-serve:v1.12.0#

ghcr.io/docling-project/docling-serve:v1.11.0#

ghcr.io/docling-project/docling-serve:v1.10.0#

ghcr.io/docling-project/docling-serve:v1.9.0#

ghcr.io/docling-project/docling-serve:v1.8.0#

ghcr.io/docling-project/docling-serve:v1.7.2#

ghcr.io/docling-project/docling-serve:v1.7.1#

ghcr.io/docling-project/docling-serve:v1.7.0#

ghcr.io/docling-project/docling-serve:v1.6.0#

ghcr.io/docling-project/docling-serve:v1.5.1#

ghcr.io/docling-project/docling-serve:v1.5.0#

ghcr.io/docling-project/docling-serve:v1.4.1#

ghcr.io/docling-project/docling-serve:v1.4.0#

ghcr.io/docling-project/docling-serve:v1.3.1#

ghcr.io/docling-project/docling-serve:v1.3.0#

ghcr.io/docling-project/docling-serve:v1.2.2#

ghcr.io/docling-project/docling-serve:v1.2.1#

ghcr.io/docling-project/docling-serve:v1.2.0#

ghcr.io/docling-project/docling-serve:v1.1.0#

ghcr.io/docling-project/docling-serve:v1.0.1#

ghcr.io/docling-project/docling-serve:v1.0.0#

The `DoclingServeApi` interface#

Requests: `ConvertDocumentRequest`#

Responses: `InBodyConvertDocumentResponse`, `PreSignedUrlConvertDocumentResponse`, `PreSignedUrlConvertResponse`, `ZipArchiveConvertDocumentResponse`, and `DocumentResponse`#