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. Its only other required dependency is SLF4J (slf4j-api), and 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:0.7.1")
}

<dependency>
  <groupId>ai.docling</groupId>
  <artifactId>docling-serve-api</artifactId>
  <version>0.7.1</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.

GraalVM Native Image

The API JAR ships reflection metadata at META-INF/native-image/ai.docling/docling-serve-api/reflect-config.json. It covers request and response models, polymorphic sources and targets, enums, Lombok builders (including inherited builders), and custom Jackson serializers and deserializers. The transitive docling-core dependency supplies metadata for DoclingDocument and its nested types.

GraalVM discovers these files automatically. Spring Boot native applications can use the models without a custom RuntimeHintsRegistrar. Include your chosen Jackson version as usual; the metadata supports Jackson 2 and Jackson 3 separately and does not add either as a runtime dependency.

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 creating instances of DoclingServeApi. An application (or downstream framework) can implement ai.docling.serve.api.spi.DoclingServeApiProvider and register it via the META-INF/services/ai.docling.serve.api.spi.DoclingServeApiProvider file, or with a provides clause in module-info.java.

DoclingServeApi.builder() collects the configuration into an immutable DoclingServeApiConfig, and build() hands it to the single available provider:

public final class MyProvider implements DoclingServeApiProvider {
  @Override
  public DoclingServeApi create(DoclingServeApiConfig config) {
    return new MyDoclingServeApi(config);
  }
}

Because providers receive a configuration object rather than implementing a builder, new options can be added in later releases without breaking existing providers. The API they create reports its configuration through DoclingServeApi.config().

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

Declaring unsupported options

A provider that cannot honor an option declares it in unsupportedOptions(), together with what should happen when a caller explicitly sets it:

• IGNORE — silently ignore the option.
• WARN — log a warning naming the option and the provider.
• FAIL — make build() throw an UnsupportedConfigurationException.

Options left at their default value are never reported. For example, a reactive implementation that manages its own threads can declare the async executor as meaningless for it:

@Override
public Map<ConfigOption<?>, Unsupported> unsupportedOptions() {
  return Map.of(DoclingServeApiConfig.ASYNC_EXECUTOR, Unsupported.WARN);
}

Options absent from this map are assumed to be honored. Providers must not silently ignore the base URL, the API key or the timeouts.

Migrating from DoclingServeApiBuilderFactory

The previous SPI, ai.docling.serve.api.spi.DoclingServeApiBuilderFactory, is deprecated for removal. It is still used when no DoclingServeApiProvider is available, but options added after its deprecation (such as asyncExecutor) make build() fail through it. See Migrating to DoclingServeApiProvider for a step-by-step guide.

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. A 422 response that carries no validation details, such as one produced by a gateway or proxy in front of docling-serve, is reported as a DoclingServeClientException, like any other error response, so that the status code and the response body remain available. This also covers a 422 body that cannot be read as JSON at all, such as an HTML or plain-text error page from a gateway: it is treated the same as a body with no validation details, not lost to a parse failure.

Logging and builders

DoclingServeApi.builder() exposes logRequests() and logResponses() for simple diagnostics. To create a modified copy of an existing API, start from its configuration:

DoclingServeApi newApi = api.config()
    .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 2026-10-05T04:35:53.055173992Z

Here are the results:

Tag	Result	Details
v1.36.0	✅ SUCCESS	Click for run details
v1.35.0	✅ SUCCESS	Click for run details
v1.34.0	✅ SUCCESS	Click for run details
v1.33.0	✅ SUCCESS	Click for run details
v1.32.0	✅ SUCCESS	Click for run 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:v1.36.0

Click to expand

###### Message

Click to collapse

Tag v1.36.0 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:   04:35:25 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:25 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:35:25 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:35:26 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:35:26 - docling_jobkit.connectors.plugins.defaults - Connector 'BoxSourceProcessor' skipped — optional dependency not installed (No module named 'box_sdk_gen'). Install the matching extra to enable it.
INFO:   04:35:26 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:35:26 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:35:26 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:35:26 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:35:26 - 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:   04:35:26 - docling_jobkit.connectors.plugins.defaults - Connector 'BoxTargetProcessor' skipped — optional dependency not installed (No module named 'box_sdk_gen'). Install the matching extra to enable it.
INFO:   04:35:26 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:35:26 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:35:26 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/readyz$,/livez$,/ready$,/healthz$,/health$)
INFO:   04:35:26 - uvicorn.error - Started server process [1]
INFO:   04:35:26 - uvicorn.error - Waiting for application startup.
INFO:   04:35:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:30 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:35:30 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:30 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:35:30 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:30 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 753fbf7f0ae7df9c46b34e2baf9cb4c2
INFO:   04:35:30 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:30 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:35:30 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:35:30.411394831 [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]+
INFO:   04:35:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:35:30,717 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:30,719 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:35:30,803 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:30,803 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:35:30,869 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:30,869 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:35:30 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:35:30 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:35:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  57%|█████▋    | 439/770 [00:00<00:00, 4388.13it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4313.23it/s]
INFO:   04:35:31 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:35:31 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:35:32 - uvicorn.error - Application startup complete.
INFO:   04:35:32 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:35:32 - docling_serve.app - Health check requested
INFO:   04:35:32 - uvicorn.access - 172.17.0.1:54662 - "GET /health HTTP/1.1" 200
INFO:   04:35:32 - docling_serve.app - Health check requested
INFO:   04:35:32 - uvicorn.access - 172.17.0.1:54674 - "GET /health HTTP/1.1" 200
INFO:   04:35:32 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:35:32 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:35:32 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:35:32 - docling_serve.app - [TENANT_ID] Task a409703a-f9d3-4000-b9a6-fe17db9646cb created with tenant_id='default'
INFO:   04:35:32 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task a409703a-f9d3-4000-b9a6-fe17db9646cb
INFO:   04:35:32 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:33 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash bd786327e14f6bf3b92eec179fb48bbb
INFO:   04:35:33 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:35:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:35:33,260 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:33,260 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:35:33,332 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:33,332 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:35:33,370 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:33,370 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:35:33 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:35:33 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:35:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  56%|█████▌    | 428/770 [00:00<00:00, 4277.98it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4210.09it/s]
INFO:   04:35:34 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:35:34 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:35:34 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:35:34 - docling.document_converter - Going to convert document batch...
INFO:   04:35:34 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash bd786327e14f6bf3b92eec179fb48bbb
INFO:   04:35:34 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:35:34 - docling.document_converter - Finished converting document file in 0.44 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:675: 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:   04:35:35 - docling_jobkit.convert.results - Processed 1 docs in 0.56 seconds.
INFO:   04:35:35 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job a409703a-f9d3-4000-b9a6-fe17db9646cb in 0.56 seconds
INFO:   04:35:36 - uvicorn.access - 172.17.0.1:54674 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:35:36 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:35:36 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:35:36 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:35:36 - docling_serve.app - [TENANT_ID] Task 7a3b3ccf-eefd-4054-a515-5c3747833c94 created with tenant_id='default'
INFO:   04:35:36 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 7a3b3ccf-eefd-4054-a515-5c3747833c94
INFO:   04:35:36 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:37 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash bd786327e14f6bf3b92eec179fb48bbb
INFO:   04:35:37 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:35:37 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:35:37,232 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:37,232 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:35:37,276 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:37,276 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:35:37,336 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:37,336 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:35:37 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:35:37 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:35:37 - 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, 4569.17it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4440.94it/s]
INFO:   04:35:37 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:35:37 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:35:38 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:35:38 - docling.document_converter - Going to convert document batch...
INFO:   04:35:38 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash bd786327e14f6bf3b92eec179fb48bbb
INFO:   04:35:38 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:35:38 - docling.document_converter - Finished converting document file in 0.38 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:675: 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:   04:35:38 - docling_jobkit.convert.results - Processed 1 docs in 0.39 seconds.
INFO:   04:35:38 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 7a3b3ccf-eefd-4054-a515-5c3747833c94 in 0.39 seconds
INFO:   04:35:38 - uvicorn.access - 172.17.0.1:54674 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:35:38 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:35:38 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:35:38 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:35:38 - docling_serve.app - [TENANT_ID] Task 8699af29-62f0-40b5-85d1-6c719953d492 created with tenant_id='default'
INFO:   04:35:38 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 8699af29-62f0-40b5-85d1-6c719953d492
INFO:   04:35:38 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:39 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 6e59dc6c44764f3927c84e26062287a0
INFO:   04:35:39 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:35:39 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:35:39,286 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:39,286 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:35:39,354 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:39,354 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:35:39,402 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:39,402 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:35:39 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:35:39 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:35:39 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  61%|██████    | 468/770 [00:00<00:00, 4676.20it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4412.11it/s]
INFO:   04:35:40 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:35:40 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:35:40 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:35:41 - docling.document_converter - Going to convert document batch...
INFO:   04:35:41 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 6e59dc6c44764f3927c84e26062287a0
INFO:   04:35:41 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:35:41 - docling.document_converter - Finished converting document file in 0.35 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:675: 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:   04:35:41 - docling_jobkit.convert.results - Processed 1 docs in 0.39 seconds.
INFO:   04:35:41 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 8699af29-62f0-40b5-85d1-6c719953d492 in 0.39 seconds
INFO:   04:35:42 - uvicorn.access - 172.17.0.1:54674 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:35:43 - uvicorn.access - 172.17.0.1:54674 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:35:43 - uvicorn.access - 172.17.0.1:54674 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.35.0

Click to expand

###### Message

Click to collapse

Tag v1.35.0 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:   04:35:11 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:11 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:35:11 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:35:13 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:35:13 - docling_jobkit.connectors.plugins.defaults - Connector 'BoxSourceProcessor' skipped — optional dependency not installed (No module named 'box_sdk_gen'). Install the matching extra to enable it.
INFO:   04:35:13 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:35:13 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:35:13 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:35:13 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:35:13 - 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:   04:35:13 - docling_jobkit.connectors.plugins.defaults - Connector 'BoxTargetProcessor' skipped — optional dependency not installed (No module named 'box_sdk_gen'). Install the matching extra to enable it.
INFO:   04:35:13 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:35:13 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:35:13 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/readyz$,/metrics$,/ready$,/health$,/livez$,/healthz$)
INFO:   04:35:14 - uvicorn.error - Started server process [1]
INFO:   04:35:14 - uvicorn.error - Waiting for application startup.
INFO:   04:35:17 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:17 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:35:17 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:17 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:35:17 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:18 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 9db6889b54cb77f8a952dbed50374a38
INFO:   04:35:18 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:35:18 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:35:18 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:35:18.153003445 [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]+
INFO:   04:35:18 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:35:18,748 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:18,750 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:35:18,884 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:18,884 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:35:18,930 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:18,930 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:35:19 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:35:19 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:35:19 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  56%|█████▌    | 428/770 [00:00<00:00, 4270.80it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4206.95it/s]
INFO:   04:35:19 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:35:19 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:35:20 - uvicorn.error - Application startup complete.
INFO:   04:35:20 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:35:21 - docling_serve.app - Health check requested
INFO:   04:35:21 - uvicorn.access - 172.17.0.1:44790 - "GET /health HTTP/1.1" 200
INFO:   04:35:21 - docling_serve.app - Health check requested
INFO:   04:35:21 - uvicorn.access - 172.17.0.1:44802 - "GET /health HTTP/1.1" 200
INFO:   04:35:21 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:35:21 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:35:21 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:35:21 - docling_serve.app - [TENANT_ID] Task c85f327e-a966-4722-8fcd-aa165d185014 created with tenant_id='default'
INFO:   04:35:21 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task c85f327e-a966-4722-8fcd-aa165d185014
INFO:   04:35:21 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:22 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash f146b2a8a5ff05cfece0b46b6075a40b
INFO:   04:35:22 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:35:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:35:22,166 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:22,166 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:35:22,216 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:22,216 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:35:22,267 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:22,267 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:35:22 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:35:22 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:35:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 475/770 [00:00<00:00, 4723.14it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4740.11it/s]
INFO:   04:35:22 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:35:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:35:23 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:35:23 - docling.document_converter - Going to convert document batch...
INFO:   04:35:23 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash f146b2a8a5ff05cfece0b46b6075a40b
INFO:   04:35:23 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:35:23 - docling.document_converter - Finished converting document file in 0.41 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:661: 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:   04:35:23 - docling_jobkit.convert.results - Processed 1 docs in 0.49 seconds.
INFO:   04:35:23 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job c85f327e-a966-4722-8fcd-aa165d185014 in 0.49 seconds
INFO:   04:35:23 - uvicorn.access - 172.17.0.1:44802 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:35:23 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:35:23 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:35:23 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:35:23 - docling_serve.app - [TENANT_ID] Task 1bea9dfd-5172-4362-8a20-2ea53109d41d created with tenant_id='default'
INFO:   04:35:23 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 1bea9dfd-5172-4362-8a20-2ea53109d41d
INFO:   04:35:23 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:24 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash f146b2a8a5ff05cfece0b46b6075a40b
INFO:   04:35:24 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:35:24 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:35:24,196 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:24,196 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:35:24,248 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:24,249 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:35:24,323 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:24,323 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:35:24 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:35:24 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:35:24 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  55%|█████▍    | 422/770 [00:00<00:00, 4211.61it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4281.13it/s]
INFO:   04:35:24 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:35:24 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:35:25 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:35:25 - docling.document_converter - Going to convert document batch...
INFO:   04:35:25 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash f146b2a8a5ff05cfece0b46b6075a40b
INFO:   04:35:25 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:35:25 - docling.document_converter - Finished converting document file in 0.74 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:661: 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:   04:35:25 - docling_jobkit.convert.results - Processed 1 docs in 0.74 seconds.
INFO:   04:35:25 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 1bea9dfd-5172-4362-8a20-2ea53109d41d in 0.74 seconds
INFO:   04:35:27 - uvicorn.access - 172.17.0.1:44802 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:35:27 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:35:27 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:35:27 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:35:27 - docling_serve.app - [TENANT_ID] Task dcef3b86-e2c1-4acf-b5fe-f708e144cf85 created with tenant_id='default'
INFO:   04:35:27 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task dcef3b86-e2c1-4acf-b5fe-f708e144cf85
INFO:   04:35:27 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:35:28 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 541ba2f8ac4b203b9e8fb2499452c6c3
INFO:   04:35:28 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:35:28 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:35:28,204 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:28,204 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:35:28,251 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:28,251 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:35:28,333 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:35:28,334 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:35:28 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:35:28 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:35:28 - 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, 4463.90it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4426.85it/s]
INFO:   04:35:28 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:35:28 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:35:29 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:35:29 - docling.document_converter - Going to convert document batch...
INFO:   04:35:29 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 541ba2f8ac4b203b9e8fb2499452c6c3
INFO:   04:35:29 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:35:29 - docling.document_converter - Finished converting document file in 0.36 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:661: 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:   04:35:29 - docling_jobkit.convert.results - Processed 1 docs in 0.39 seconds.
INFO:   04:35:29 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job dcef3b86-e2c1-4acf-b5fe-f708e144cf85 in 0.39 seconds
INFO:   04:35:29 - uvicorn.access - 172.17.0.1:44802 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:35:30 - uvicorn.access - 172.17.0.1:44802 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:35:30 - uvicorn.access - 172.17.0.1:44802 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.34.0

Click to expand

###### Message

Click to collapse

Tag v1.34.0 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:   04:32:24 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:32:24 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:32:24 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:32:25 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:32:25 - docling_jobkit.connectors.plugins.defaults - Connector 'BoxSourceProcessor' skipped — optional dependency not installed (No module named 'box_sdk_gen'). Install the matching extra to enable it.
INFO:   04:32:25 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:32:25 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:32:25 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:32:25 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:32:25 - 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:   04:32:25 - docling_jobkit.connectors.plugins.defaults - Connector 'BoxTargetProcessor' skipped — optional dependency not installed (No module named 'box_sdk_gen'). Install the matching extra to enable it.
INFO:   04:32:25 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:32:25 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:32:25 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/livez$,/healthz$,/health$,/readyz$,/ready$,/metrics$)
INFO:   04:32:25 - uvicorn.error - Started server process [1]
INFO:   04:32:25 - uvicorn.error - Waiting for application startup.
INFO:   04:32:28 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:32:28 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:32:28 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:32:28 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:32:28 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:29 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 4b071d1415ad1ebea85cea80f4d6fa7b
INFO:   04:32:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:32:29 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:32:29 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:32:29.236055798 [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]+
INFO:   04:32:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:32:29,807 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:29,810 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:32:29,892 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:29,893 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:32:29,934 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:29,935 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:32:30 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:30 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:32:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  56%|█████▌    | 432/770 [00:00<00:00, 4307.93it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4322.30it/s]
INFO:   04:32:30 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:32:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:31 - uvicorn.error - Application startup complete.
INFO:   04:32:31 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:32:31 - docling_serve.app - Health check requested
INFO:   04:32:31 - uvicorn.access - 172.17.0.1:40344 - "GET /health HTTP/1.1" 200
INFO:   04:32:31 - docling_serve.app - Health check requested
INFO:   04:32:31 - uvicorn.access - 172.17.0.1:40356 - "GET /health HTTP/1.1" 200
INFO:   04:32:31 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:32:31 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:32:31 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:32:31 - docling_serve.app - [TENANT_ID] Task 97fbe823-de21-4d93-9210-ee28d16f8d5a created with tenant_id='default'
INFO:   04:32:31 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 97fbe823-de21-4d93-9210-ee28d16f8d5a
INFO:   04:32:31 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:31 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a2e22ed4b71003a02ccc48074bc1ab74
INFO:   04:32:31 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:32:31 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:32:31,944 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:31,944 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:32:32,021 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:32,022 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:32:32,098 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:32,098 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:32:32 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:32 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:32:32 - 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, 4243.73it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4293.15it/s]
INFO:   04:32:32 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:32:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:33 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:32:33 - docling.document_converter - Going to convert document batch...
INFO:   04:32:33 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash a2e22ed4b71003a02ccc48074bc1ab74
INFO:   04:32:33 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:32:33 - docling.document_converter - Finished converting document file in 0.44 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:661: 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:    04:32:33 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:32:33 - docling_jobkit.convert.results - Processed 1 docs in 0.53 seconds.
INFO:   04:32:33 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 97fbe823-de21-4d93-9210-ee28d16f8d5a in 0.53 seconds
INFO:   04:32:33 - uvicorn.access - 172.17.0.1:40356 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:32:33 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:32:33 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:32:33 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:32:33 - docling_serve.app - [TENANT_ID] Task 09cda21f-8a4b-44f4-ba65-eaec28bb5632 created with tenant_id='default'
INFO:   04:32:33 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 09cda21f-8a4b-44f4-ba65-eaec28bb5632
INFO:   04:32:33 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:33 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a2e22ed4b71003a02ccc48074bc1ab74
INFO:   04:32:33 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:32:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:32:33,966 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:33,966 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:32:34,015 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:34,015 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:32:34,102 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:34,103 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:32:34 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:34 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:32:34 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  55%|█████▌    | 427/770 [00:00<00:00, 4266.87it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4170.52it/s]
INFO:   04:32:34 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:32:34 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:35 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:32:35 - docling.document_converter - Going to convert document batch...
INFO:   04:32:35 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash a2e22ed4b71003a02ccc48074bc1ab74
INFO:   04:32:35 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:32:35 - docling.document_converter - Finished converting document file in 0.33 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:661: 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:   04:32:35 - docling_jobkit.convert.results - Processed 1 docs in 0.33 seconds.
INFO:   04:32:35 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 09cda21f-8a4b-44f4-ba65-eaec28bb5632 in 0.33 seconds
INFO:   04:32:35 - uvicorn.access - 172.17.0.1:40356 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:32:35 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:32:35 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:32:35 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:32:35 - docling_serve.app - [TENANT_ID] Task a1049955-2d2d-45b7-89c5-de2d829e4c59 created with tenant_id='default'
INFO:   04:32:35 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task a1049955-2d2d-45b7-89c5-de2d829e4c59
INFO:   04:32:35 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:36 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 78cb5ade368814d0813875c101ee149b
INFO:   04:32:36 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:32:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:32:36,405 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:36,405 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:32:36,468 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:36,468 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:32:36,510 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:36,510 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:32:36 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:36 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:32:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  58%|█████▊    | 449/770 [00:00<00:00, 4484.67it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4386.60it/s]
INFO:   04:32:37 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:32:37 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:37 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:32:37 - docling.document_converter - Going to convert document batch...
INFO:   04:32:37 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 78cb5ade368814d0813875c101ee149b
INFO:   04:32:37 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:32:37 - docling.document_converter - Finished converting document file in 0.30 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:661: 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:   04:32:37 - docling_jobkit.convert.results - Processed 1 docs in 0.35 seconds.
INFO:   04:32:37 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job a1049955-2d2d-45b7-89c5-de2d829e4c59 in 0.35 seconds
INFO:   04:32:37 - uvicorn.access - 172.17.0.1:40356 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:32:38 - uvicorn.access - 172.17.0.1:40356 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:32:38 - uvicorn.access - 172.17.0.1:40356 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.33.0

Click to expand

###### Message

Click to collapse

Tag v1.33.0 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:   04:32:15 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:32:15 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:32:15 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:32:15 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:32:15 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:32:15 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:32:15 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:32:15 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:32:15 - 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:   04:32:15 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:32:15 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:32:15 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/readyz$,/livez$,/ready$,/health$,/healthz$,/metrics$)
INFO:   04:32:16 - uvicorn.error - Started server process [1]
INFO:   04:32:16 - uvicorn.error - Waiting for application startup.
INFO:   04:32:19 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:32:19 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:32:19 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:32:19 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:32:19 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:19 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 4b071d1415ad1ebea85cea80f4d6fa7b
INFO:   04:32:19 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:32:19 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:32:19 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:32:19.884307149 [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]+
INFO:   04:32:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:32:20,137 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:20,139 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:32:20,236 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:20,236 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:32:20,280 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:20,280 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:32:20 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:20 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:32:20 - 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, 4730.00it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4735.69it/s]
INFO:   04:32:21 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:32:21 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:21 - uvicorn.error - Application startup complete.
INFO:   04:32:21 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:32:21 - docling_serve.app - Health check requested
INFO:   04:32:21 - uvicorn.access - 172.17.0.1:55246 - "GET /health HTTP/1.1" 200
INFO:   04:32:21 - docling_serve.app - Health check requested
INFO:   04:32:21 - uvicorn.access - 172.17.0.1:55248 - "GET /health HTTP/1.1" 200
INFO:   04:32:21 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:32:21 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:32:21 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:32:21 - docling_serve.app - [TENANT_ID] Task d25754ea-f721-4faf-b5cc-f6cc97ce7dd2 created with tenant_id='default'
INFO:   04:32:21 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task d25754ea-f721-4faf-b5cc-f6cc97ce7dd2
INFO:   04:32:21 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:21 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a2e22ed4b71003a02ccc48074bc1ab74
INFO:   04:32:21 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:32:21 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:32:21,897 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:21,897 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:32:21,951 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:21,951 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:32:22,032 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:22,033 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:32:22 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:22 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:32:22 - 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, 5056.22it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4911.12it/s]
INFO:   04:32:22 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:32:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:23 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:32:23 - docling.document_converter - Going to convert document batch...
INFO:   04:32:23 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash a2e22ed4b71003a02ccc48074bc1ab74
INFO:   04:32:23 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:32:23 - docling.document_converter - Finished converting document file in 0.42 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:655: 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:    04:32:23 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:32:23 - docling_jobkit.convert.results - Processed 1 docs in 0.49 seconds.
INFO:   04:32:23 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job d25754ea-f721-4faf-b5cc-f6cc97ce7dd2 in 0.49 seconds
INFO:   04:32:23 - uvicorn.access - 172.17.0.1:55248 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:32:23 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:32:23 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:32:23 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:32:23 - docling_serve.app - [TENANT_ID] Task 2a24172f-39ce-46c2-b024-62ef451943f3 created with tenant_id='default'
INFO:   04:32:23 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 2a24172f-39ce-46c2-b024-62ef451943f3
INFO:   04:32:23 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:23 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a2e22ed4b71003a02ccc48074bc1ab74
INFO:   04:32:23 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:32:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:32:23,911 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:23,911 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:32:23,965 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:23,965 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:32:24,046 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:24,046 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:32:24 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:24 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:32:24 - 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, 5075.06it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4932.96it/s]
INFO:   04:32:24 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:32:24 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:25 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:32:25 - docling.document_converter - Going to convert document batch...
INFO:   04:32:25 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash a2e22ed4b71003a02ccc48074bc1ab74
INFO:   04:32:25 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:32:25 - docling.document_converter - Finished converting document file in 0.46 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:655: 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:   04:32:25 - docling_jobkit.convert.results - Processed 1 docs in 0.46 seconds.
INFO:   04:32:25 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 2a24172f-39ce-46c2-b024-62ef451943f3 in 0.46 seconds
INFO:   04:32:25 - uvicorn.access - 172.17.0.1:55248 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:32:25 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:32:25 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:32:25 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:32:25 - docling_serve.app - [TENANT_ID] Task 403259f9-1f84-4457-b42e-1fc881614e42 created with tenant_id='default'
INFO:   04:32:25 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 403259f9-1f84-4457-b42e-1fc881614e42
INFO:   04:32:25 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:32:25 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 78cb5ade368814d0813875c101ee149b
INFO:   04:32:25 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:32:25 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:32:25,946 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:25,946 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:32:25,992 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:25,992 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:32:26,045 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:32:26,046 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:32:26 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:32:26 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:32:26 - 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, 4656.43it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4805.56it/s]
INFO:   04:32:26 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:32:26 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:32:27 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:32:27 - docling.document_converter - Going to convert document batch...
INFO:   04:32:27 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 78cb5ade368814d0813875c101ee149b
INFO:   04:32:27 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:32:27 - docling.document_converter - Finished converting document file in 0.37 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:655: 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:   04:32:27 - docling_jobkit.convert.results - Processed 1 docs in 0.41 seconds.
INFO:   04:32:27 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 403259f9-1f84-4457-b42e-1fc881614e42 in 0.41 seconds
INFO:   04:32:27 - uvicorn.access - 172.17.0.1:55248 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:32:28 - uvicorn.access - 172.17.0.1:55248 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:32:28 - uvicorn.access - 172.17.0.1:55248 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

ghcr.io/docling-project/docling-serve:v1.32.0

Click to expand

###### Message

Click to collapse

Tag v1.32.0 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:   04:29:14 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:14 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:29:14 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:29:15 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:29:15 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:29:15 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:29:15 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:29:15 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:29:15 - 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:   04:29:15 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:29:15 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:29:15 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/health$,/ready$,/metrics$,/readyz$,/livez$,/healthz$)
INFO:   04:29:15 - uvicorn.error - Started server process [1]
INFO:   04:29:15 - uvicorn.error - Waiting for application startup.
INFO:   04:29:19 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:19 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:29:19 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:19 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
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 019ac48c98c19d19127e62d9a66de678
INFO:   04:29:19 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:19 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:29:19 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:29:19.716502065 [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]+
INFO:   04:29:19 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:29:19,970 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:19,971 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:29:20,049 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:20,050 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:29:20,091 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:20,091 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:29:20 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:20 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:29:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  69%|██████▉   | 533/770 [00:00<00:00, 5322.38it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4970.40it/s]
INFO:   04:29:21 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:29:21 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:21 - uvicorn.error - Application startup complete.
INFO:   04:29:21 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:29:22 - docling_serve.app - Health check requested
INFO:   04:29:22 - uvicorn.access - 172.17.0.1:44050 - "GET /health HTTP/1.1" 200
INFO:   04:29:22 - docling_serve.app - Health check requested
INFO:   04:29:22 - uvicorn.access - 172.17.0.1:44054 - "GET /health HTTP/1.1" 200
INFO:   04:29:22 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:22 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:22 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:22 - docling_serve.app - [TENANT_ID] Task 9d05cf72-8693-4a63-b326-a0056938ae24 created with tenant_id='default'
INFO:   04:29:22 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 9d05cf72-8693-4a63-b326-a0056938ae24
INFO:   04:29:22 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:22 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 6b5aa9d49bf5933a22591c3559b94bfd
INFO:   04:29:22 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:29:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:29:22,399 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:22,399 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:29:22,481 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:22,482 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:29:22,540 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:22,541 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:29:22 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:22 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:29:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  56%|█████▌    | 430/770 [00:00<00:00, 4265.80it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4301.58it/s]
INFO:   04:29:23 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:29:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:24 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:29:24 - docling.document_converter - Going to convert document batch...
INFO:   04:29:24 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 6b5aa9d49bf5933a22591c3559b94bfd
INFO:   04:29:24 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:24 - docling.document_converter - Finished converting document file in 0.56 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:645: 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:    04:29:24 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:29:24 - docling_jobkit.convert.results - Processed 1 docs in 0.64 seconds.
INFO:   04:29:24 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 9d05cf72-8693-4a63-b326-a0056938ae24 in 0.64 seconds
INFO:   04:29:24 - uvicorn.access - 172.17.0.1:44054 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:24 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:24 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:24 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:24 - docling_serve.app - [TENANT_ID] Task 7bc16e93-950a-4178-b2db-8de5fe154df9 created with tenant_id='default'
INFO:   04:29:24 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 7bc16e93-950a-4178-b2db-8de5fe154df9
INFO:   04:29:24 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:24 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 6b5aa9d49bf5933a22591c3559b94bfd
INFO:   04:29:24 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:29:24 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:29:24,426 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:24,427 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:29:24,489 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:24,490 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:29:24,534 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:24,534 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:29:24 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:24 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:29:24 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  66%|██████▌   | 505/770 [00:00<00:00, 5048.79it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5013.26it/s]
INFO:   04:29:25 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:29:25 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:25 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:29:25 - docling.document_converter - Going to convert document batch...
INFO:   04:29:25 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 6b5aa9d49bf5933a22591c3559b94bfd
INFO:   04:29:25 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:25 - docling.document_converter - Finished converting document file in 0.39 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:645: 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:   04:29:25 - docling_jobkit.convert.results - Processed 1 docs in 0.40 seconds.
INFO:   04:29:25 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 7bc16e93-950a-4178-b2db-8de5fe154df9 in 0.40 seconds
INFO:   04:29:26 - uvicorn.access - 172.17.0.1:44054 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:26 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:26 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:26 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:26 - docling_serve.app - [TENANT_ID] Task be4811f6-ea02-4fb1-b9aa-80bf79fa8746 created with tenant_id='default'
INFO:   04:29:26 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task be4811f6-ea02-4fb1-b9aa-80bf79fa8746
INFO:   04:29:26 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:26 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 345bf8d309571831952bb6d347d0f005
INFO:   04:29:26 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:29:26 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:29:26,455 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:26,455 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:29:26,500 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:26,501 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:29:26,579 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:26,579 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:29:26 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:26 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:29:26 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  53%|█████▎    | 410/770 [00:00<00:00, 4083.98it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4021.25it/s]
INFO:   04:29:27 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:29:27 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:27 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:29:27 - docling.document_converter - Going to convert document batch...
INFO:   04:29:27 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 345bf8d309571831952bb6d347d0f005
INFO:   04:29:27 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:27 - docling.document_converter - Finished converting document file in 0.29 sec.
/opt/app-root/lib64/python3.12/site-packages/docling/datamodel/base_models.py:645: 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:   04:29:27 - docling_jobkit.convert.results - Processed 1 docs in 0.33 seconds.
INFO:   04:29:27 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job be4811f6-ea02-4fb1-b9aa-80bf79fa8746 in 0.33 seconds
INFO:   04:29:28 - uvicorn.access - 172.17.0.1:44054 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:28 - uvicorn.access - 172.17.0.1:44054 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:29:28 - uvicorn.access - 172.17.0.1:44054 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200

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:   04:29:10 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:10 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:29:10 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:29:11 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:29:11 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:29:11 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:29:11 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:29:11 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:29:11 - 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:   04:29:11 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:29:11 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:29:11 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/readyz$,/livez$,/health$,/ready$,/healthz$)
INFO:   04:29:11 - uvicorn.error - Started server process [1]
INFO:   04:29:11 - uvicorn.error - Waiting for application startup.
INFO:   04:29:14 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:15 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:29:15 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:15 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
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:   04:29:15 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 9d06178a9f1c8aeb9889d1e46440f3fa
INFO:   04:29:15 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:29:15 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:29:15 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:29:15.406423661 [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]+
INFO:   04:29:15 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:29:15,654 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:15,656 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:29:15,758 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:15,758 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:29:15,822 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:15,823 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:29:15 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:15 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:29:15 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▊    | 451/770 [00:00<00:00, 4475.04it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4434.43it/s]
INFO:   04:29:16 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:29:16 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:17 - uvicorn.error - Application startup complete.
INFO:   04:29:17 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:29:18 - docling_serve.app - Health check requested
INFO:   04:29:18 - uvicorn.access - 172.17.0.1:53048 - "GET /health HTTP/1.1" 200
INFO:   04:29:18 - docling_serve.app - Health check requested
INFO:   04:29:18 - uvicorn.access - 172.17.0.1:53064 - "GET /health HTTP/1.1" 200
INFO:   04:29:18 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:18 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:18 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:18 - docling_serve.app - [TENANT_ID] Task 55492550-f691-45d2-916d-572d2893d0d4 created with tenant_id='default'
INFO:   04:29:18 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 55492550-f691-45d2-916d-572d2893d0d4
INFO:   04:29:18 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:18 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a1d02772fb798c5f3e06aa82e0d32e7d
INFO:   04:29:18 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:29:18 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:29:18,395 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:18,395 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:29:18,454 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:18,454 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:29:18,535 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:18,536 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:29:18 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:18 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:29:18 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▊    | 452/770 [00:00<00:00, 4516.43it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4472.39it/s]
INFO:   04:29:19 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:29:19 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:19 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:29:19 - docling.document_converter - Going to convert document batch...
INFO:   04:29:19 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash a1d02772fb798c5f3e06aa82e0d32e7d
INFO:   04:29:19 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:19 - 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:    04:29:19 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:29:19 - docling_jobkit.convert.results - Processed 1 docs in 0.47 seconds.
INFO:   04:29:19 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 55492550-f691-45d2-916d-572d2893d0d4 in 0.47 seconds
INFO:   04:29:20 - uvicorn.access - 172.17.0.1:53064 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:20 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:20 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:20 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:20 - docling_serve.app - [TENANT_ID] Task 15550fbb-b592-4e2a-b356-274ace4b3167 created with tenant_id='default'
INFO:   04:29:20 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 15550fbb-b592-4e2a-b356-274ace4b3167
INFO:   04:29:20 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:20 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a1d02772fb798c5f3e06aa82e0d32e7d
INFO:   04:29:20 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:29:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:29:20,467 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:20,467 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:29:20,520 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:20,520 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:29:20,607 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:20,607 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:29:20 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:20 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:29:20 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  79%|███████▉  | 607/770 [00:00<00:00, 6068.13it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 7138.91it/s]
INFO:   04:29:21 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:29:21 - 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 a1d02772fb798c5f3e06aa82e0d32e7d
INFO:   04:29:21 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:21 - docling.document_converter - Finished converting document file in 0.34 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:   04:29:21 - docling_jobkit.convert.results - Processed 1 docs in 0.34 seconds.
INFO:   04:29:21 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 15550fbb-b592-4e2a-b356-274ace4b3167 in 0.34 seconds
INFO:   04:29:22 - uvicorn.access - 172.17.0.1:53064 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:22 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:29:22 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:29:22 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:29:22 - docling_serve.app - [TENANT_ID] Task f71ff45e-fe71-4231-b844-b78f175fb9f6 created with tenant_id='default'
INFO:   04:29:22 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task f71ff45e-fe71-4231-b844-b78f175fb9f6
INFO:   04:29:22 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:29:22 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 21c1cdafbd5b48f9b0cfcb806e9e9d57
INFO:   04:29:22 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:29:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:29:22,436 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:22,437 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:29:22,577 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:22,577 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:29:22,664 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:29:22,665 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:29:22 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:29:22 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:29:22 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  56%|█████▌    | 433/770 [00:00<00:00, 4326.28it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4429.46it/s]
INFO:   04:29:23 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:29:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:29:23 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:29:23 - docling.document_converter - Going to convert document batch...
INFO:   04:29:23 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 21c1cdafbd5b48f9b0cfcb806e9e9d57
INFO:   04:29:23 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:29:23 - 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:   04:29:23 - docling_jobkit.convert.results - Processed 1 docs in 0.40 seconds.
INFO:   04:29:23 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job f71ff45e-fe71-4231-b844-b78f175fb9f6 in 0.40 seconds
INFO:   04:29:24 - uvicorn.access - 172.17.0.1:53064 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:29:24 - uvicorn.access - 172.17.0.1:53064 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:29:24 - uvicorn.access - 172.17.0.1:53064 - "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:   04:26:22 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:22 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:26:22 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:26:22 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointSourceProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:26:22 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:26:22 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:26:22 - docling_jobkit.connectors.plugins.defaults - Connector 'AstraDBTargetProcessor' skipped — optional dependency not installed (No module named 'astrapy'). Install the matching extra to enable it.
INFO:   04:26:22 - docling_jobkit.connectors.plugins.defaults - Connector 'SharePointTargetProcessor' skipped — optional dependency not installed (No module named 'office365'). Install the matching extra to enable it.
INFO:   04:26:22 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:26:22 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:26:22 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/livez$,/metrics$,/ready$,/healthz$,/readyz$,/health$)
INFO:   04:26:23 - uvicorn.error - Started server process [1]
INFO:   04:26:23 - uvicorn.error - Waiting for application startup.
INFO:   04:26:26 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:26 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:26:26 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:26 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:26:26 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:26 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash e2db689ab35ca52a08b5b7fa25de8cc8
INFO:   04:26:26 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:26 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:26:26 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:26:26.833886335 [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]+
INFO:   04:26:27 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:26:27,060 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:27,061 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:26:27,115 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:27,115 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:26:27,153 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:27,154 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:26:27 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:27 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:26:27 - 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:  66%|██████▌   | 508/770 [00:00<00:00, 5045.18it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5093.92it/s]
INFO:   04:26:29 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:26:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:29 - uvicorn.error - Application startup complete.
INFO:   04:26:29 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:26:30 - docling_serve.app - Health check requested
INFO:   04:26:30 - uvicorn.access - 172.17.0.1:39942 - "GET /health HTTP/1.1" 200
INFO:   04:26:30 - docling_serve.app - Health check requested
INFO:   04:26:30 - uvicorn.access - 172.17.0.1:39952 - "GET /health HTTP/1.1" 200
INFO:   04:26:30 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:26:30 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:26:30 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:26:30 - docling_serve.app - [TENANT_ID] Task f9999c15-e4e5-4875-897e-e49884a763aa created with tenant_id='default'
INFO:   04:26:30 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task f9999c15-e4e5-4875-897e-e49884a763aa
INFO:   04:26:30 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:30 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 5cd3e3811d2e4d7d996fcfcbbe327b1d
INFO:   04:26:30 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:26:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:26:30,545 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:30,545 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:26:30,605 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:30,605 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:26:30,647 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:30,647 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:26:30 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:30 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:26:30 - 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, 4687.12it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4962.04it/s]
INFO:   04:26:31 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:26:31 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:31 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:31 - docling.document_converter - Going to convert document batch...
INFO:   04:26:31 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 5cd3e3811d2e4d7d996fcfcbbe327b1d
INFO:   04:26:31 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:31 - docling.document_converter - Finished converting document file in 0.43 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:    04:26:31 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:26:32 - docling_jobkit.convert.results - Processed 1 docs in 0.52 seconds.
INFO:   04:26:32 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job f9999c15-e4e5-4875-897e-e49884a763aa in 0.52 seconds
INFO:   04:26:32 - uvicorn.access - 172.17.0.1:39952 - "POST /v1/convert/source 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 833747c1-c3da-4a63-aee2-9c396f724eba created with tenant_id='default'
INFO:   04:26:32 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 833747c1-c3da-4a63-aee2-9c396f724eba
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 5cd3e3811d2e4d7d996fcfcbbe327b1d
INFO:   04:26:32 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:26:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:26:32,561 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:32,561 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:26:32,628 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:32,628 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:26:32,668 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:32,668 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:26:32 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:32 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:26:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  66%|██████▋   | 511/770 [00:00<00:00, 5107.78it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5051.14it/s]
INFO:   04:26:33 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:26:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:33 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:34 - docling.document_converter - Going to convert document batch...
INFO:   04:26:34 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 5cd3e3811d2e4d7d996fcfcbbe327b1d
INFO:   04:26:34 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:34 - docling.document_converter - Finished converting document file in 0.72 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:   04:26:34 - docling_jobkit.convert.results - Processed 1 docs in 0.72 seconds.
INFO:   04:26:34 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 833747c1-c3da-4a63-aee2-9c396f724eba in 0.72 seconds
INFO:   04:26:34 - uvicorn.access - 172.17.0.1:39952 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:26:34 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:26:34 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:26:34 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:26:34 - docling_serve.app - [TENANT_ID] Task af2dff63-0b10-4ff8-a6eb-d4f7eaa8523a created with tenant_id='default'
INFO:   04:26:34 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task af2dff63-0b10-4ff8-a6eb-d4f7eaa8523a
INFO:   04:26:34 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:34 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 9df8df3fc03c33cf6e8bd583e67dce64
INFO:   04:26:34 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:26:34 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:26:34,613 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:34,613 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:26:34,702 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:34,702 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/ch_ppocr_mobile_v2.0_cls_mobile.onnx
[INFO] 2026-10-05 04:26:34,779 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:34,779 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/PP-OCRv6_rec_small.onnx
INFO:   04:26:34 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:34 - docling.models.inference_engines.object_detection.transformers_engine - Initializing Transformers object-detection engine
INFO:   04:26:34 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  58%|█████▊    | 450/770 [00:00<00:00, 4497.59it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4420.55it/s]
INFO:   04:26:35 - docling.models.inference_engines.object_detection.transformers_engine - Transformers engine ready (device=cpu, dtype=torch.float32)
INFO:   04:26:35 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:36 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:36 - docling.document_converter - Going to convert document batch...
INFO:   04:26:36 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 9df8df3fc03c33cf6e8bd583e67dce64
INFO:   04:26:36 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:36 - docling.document_converter - Finished converting document file in 0.59 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:   04:26:36 - docling_jobkit.convert.results - Processed 1 docs in 0.64 seconds.
INFO:   04:26:36 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job af2dff63-0b10-4ff8-a6eb-d4f7eaa8523a in 0.64 seconds
INFO:   04:26:36 - uvicorn.access - 172.17.0.1:39952 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:26:36 - uvicorn.access - 172.17.0.1:39952 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:26:36 - uvicorn.access - 172.17.0.1:39952 - "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:   04:26:18 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:18 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:26:18 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:26:19 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:26:19 - docling_jobkit.connectors.plugins.defaults - Connector 'OpenSearchTargetProcessor' skipped — optional dependency not installed (No module named 'opensearchpy'). Install the matching extra to enable it.
INFO:   04:26:19 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:26:19 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:26:19 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/livez$,/healthz$,/readyz$,/metrics$,/ready$,/health$)
INFO:   04:26:19 - uvicorn.error - Started server process [1]
INFO:   04:26:19 - uvicorn.error - Waiting for application startup.
INFO:   04:26:22 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:22 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:26:22 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:22 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:26:22 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:22 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash caff667a02b7f037838d32d34acfe0b0
INFO:   04:26:22 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:26:22 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:26:22 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:26:22.817763931 [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]+
INFO:   04:26:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:26:23,067 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:23,068 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:26:23,168 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:23,168 [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] 2026-10-05 04:26:23,209 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:23,209 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:26:23 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:23 - 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, 5007.63it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5105.93it/s]
INFO:   04:26:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:24 - uvicorn.error - Application startup complete.
INFO:   04:26:24 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:26:25 - docling_serve.app - Health check requested
INFO:   04:26:25 - uvicorn.access - 172.17.0.1:34802 - "GET /health HTTP/1.1" 200
INFO:   04:26:25 - docling_serve.app - Health check requested
INFO:   04:26:25 - uvicorn.access - 172.17.0.1:34804 - "GET /health HTTP/1.1" 200
INFO:   04:26:25 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:26:25 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:26:25 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:26:25 - docling_serve.app - [TENANT_ID] Task ba64a8d4-dd35-4e8f-913c-1bfb23d21960 created with tenant_id='default'
INFO:   04:26:25 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task ba64a8d4-dd35-4e8f-913c-1bfb23d21960
INFO:   04:26:25 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:25 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash ab12ef348c8fc273bf2be7a49773b303
INFO:   04:26:25 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:26:25 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:26:25,516 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:25,517 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:26:25,565 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:25,566 [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] 2026-10-05 04:26:25,650 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:25,650 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:26:25 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:25 - 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, 4669.00it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4724.57it/s]
INFO:   04:26:26 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:27 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:27 - docling.document_converter - Going to convert document batch...
INFO:   04:26:27 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:26:27 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:27 - docling.document_converter - Finished converting document file in 0.53 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:    04:26:27 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:26:27 - docling_jobkit.convert.results - Processed 1 docs in 0.61 seconds.
INFO:   04:26:27 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job ba64a8d4-dd35-4e8f-913c-1bfb23d21960 in 0.61 seconds
INFO:   04:26:29 - uvicorn.access - 172.17.0.1:34804 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:26:29 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:26:29 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:26:29 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:26:29 - docling_serve.app - [TENANT_ID] Task 2e92ba1c-fe7b-4c47-ba40-3d463f20de57 created with tenant_id='default'
INFO:   04:26:29 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 2e92ba1c-fe7b-4c47-ba40-3d463f20de57
INFO:   04:26:29 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:29 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash ab12ef348c8fc273bf2be7a49773b303
INFO:   04:26:29 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:26:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:26:29,569 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:29,569 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:26:29,620 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:29,621 [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] 2026-10-05 04:26:29,664 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:29,664 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.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:  70%|██████▉   | 536/770 [00:00<00:00, 5356.95it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5417.22it/s]
INFO:   04:26:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:30 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:30 - docling.document_converter - Going to convert document batch...
INFO:   04:26:30 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:26:30 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:31 - 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(
INFO:   04:26:31 - docling_jobkit.convert.results - Processed 1 docs in 0.60 seconds.
INFO:   04:26:31 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 2e92ba1c-fe7b-4c47-ba40-3d463f20de57 in 0.60 seconds
INFO:   04:26:31 - uvicorn.access - 172.17.0.1:34804 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:26:31 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:26:31 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:26:31 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:26:31 - docling_serve.app - [TENANT_ID] Task 17c12ece-e86b-4987-b6d2-8aa842fff61d created with tenant_id='default'
INFO:   04:26:31 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 17c12ece-e86b-4987-b6d2-8aa842fff61d
INFO:   04:26:31 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:26:31 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash b3849eabdc7ecae5f3d54428a0162342
INFO:   04:26:31 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:26:31 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:26:31,592 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:31,593 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:26:31,635 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:31,636 [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] 2026-10-05 04:26:31,676 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:26:31,677 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:26:31 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:26:31 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 516/770 [00:00<00:00, 5131.56it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5205.82it/s]
INFO:   04:26:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:26:32 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:26:32 - docling.document_converter - Going to convert document batch...
INFO:   04:26:32 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:26:32 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:26:32 - docling.document_converter - Finished converting document file in 0.54 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:   04:26:33 - docling_jobkit.convert.results - Processed 1 docs in 0.61 seconds.
INFO:   04:26:33 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 17c12ece-e86b-4987-b6d2-8aa842fff61d in 0.61 seconds
INFO:   04:26:33 - uvicorn.access - 172.17.0.1:34804 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:26:33 - uvicorn.access - 172.17.0.1:34804 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:26:33 - uvicorn.access - 172.17.0.1:34804 - "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:   04:23:25 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:25 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:23:25 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:23:26 - docling_jobkit.connectors.connector_factory - Loading connector plugin 'docling_jobkit_defaults'
INFO:   04:23:26 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:23:26 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:23:26 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/ready$,/metrics$,/livez$,/healthz$,/readyz$,/health$)
INFO:   04:23:26 - uvicorn.error - Started server process [1]
INFO:   04:23:26 - uvicorn.error - Waiting for application startup.
INFO:   04:23:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:29 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:23:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:29 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:23:29 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:29 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 118619a1d3ed3b201b814dac997d6742
INFO:   04:23:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:29 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:23:29 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:23:29.714420013 [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]+
INFO:   04:23:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:23:29,981 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:29,983 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:23:30,055 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:30,058 [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] 2026-10-05 04:23:30,126 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:30,126 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:23:30 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:30 - 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, 4635.97it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 6436.07it/s]
INFO:   04:23:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:31 - uvicorn.error - Application startup complete.
INFO:   04:23:31 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:23:32 - docling_serve.app - Health check requested
INFO:   04:23:32 - uvicorn.access - 172.17.0.1:58272 - "GET /health HTTP/1.1" 200
INFO:   04:23:32 - docling_serve.app - Health check requested
INFO:   04:23:32 - uvicorn.access - 172.17.0.1:58286 - "GET /health HTTP/1.1" 200
INFO:   04:23:32 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:23:32 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:23:32 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:23:32 - docling_serve.app - [TENANT_ID] Task d53e5806-60b3-4d61-9e29-a9a4ea32439e created with tenant_id='default'
INFO:   04:23:32 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task d53e5806-60b3-4d61-9e29-a9a4ea32439e
INFO:   04:23:32 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:32 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 83f6e7119a1499def48d21cd30b03869
INFO:   04:23:32 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:23:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:23:32,676 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:32,676 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:23:32,755 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:32,755 [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] 2026-10-05 04:23:32,805 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:32,805 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:23:32 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  69%|██████▉   | 534/770 [00:00<00:00, 5322.58it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5285.13it/s]
INFO:   04:23:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:34 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:23:34 - docling.document_converter - Going to convert document batch...
INFO:   04:23:34 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:23:34 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:23:34 - docling.document_converter - Finished converting document file in 0.45 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:    04:23:34 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:23:34 - docling_jobkit.convert.results - Processed 1 docs in 0.56 seconds.
INFO:   04:23:34 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job d53e5806-60b3-4d61-9e29-a9a4ea32439e in 0.56 seconds
INFO:   04:23:36 - uvicorn.access - 172.17.0.1:58286 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:23:36 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:23:36 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:23:36 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:23:36 - docling_serve.app - [TENANT_ID] Task b59bdd1a-719b-41d5-a3cd-8123029e1793 created with tenant_id='default'
INFO:   04:23:36 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task b59bdd1a-719b-41d5-a3cd-8123029e1793
INFO:   04:23:36 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:36 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 83f6e7119a1499def48d21cd30b03869
INFO:   04:23:36 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:23:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:23:36,690 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:36,690 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:23:36,744 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:36,744 [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] 2026-10-05 04:23:36,811 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:36,812 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:23:36 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:36 - 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, 5265.64it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5287.04it/s]
INFO:   04:23:37 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:37 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:23:37 - docling.document_converter - Going to convert document batch...
INFO:   04:23:37 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:23:37 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:23:37 - docling.document_converter - Finished converting document file in 0.36 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:   04:23:37 - docling_jobkit.convert.results - Processed 1 docs in 0.36 seconds.
INFO:   04:23:37 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job b59bdd1a-719b-41d5-a3cd-8123029e1793 in 0.36 seconds
INFO:   04:23:38 - uvicorn.access - 172.17.0.1:58286 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:23:38 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:23:38 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:23:38 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:23:38 - docling_serve.app - [TENANT_ID] Task 3e0195c3-2a0c-493f-ba94-832693d7a6b3 created with tenant_id='default'
INFO:   04:23:38 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 3e0195c3-2a0c-493f-ba94-832693d7a6b3
INFO:   04:23:38 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:38 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 528ff53b5cc66cd5f7f61ef2cd72a8fc
INFO:   04:23:38 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:23:38 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:23:38,750 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:38,750 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:23:38,799 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:38,800 [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] 2026-10-05 04:23:38,877 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:38,877 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:23:38 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:38 - 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, 4576.39it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4626.21it/s]
INFO:   04:23:39 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:39 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:23:40 - docling.document_converter - Going to convert document batch...
INFO:   04:23:40 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:23:40 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:23:40 - docling.document_converter - Finished converting document file in 0.40 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:   04:23:40 - docling_jobkit.convert.results - Processed 1 docs in 0.45 seconds.
INFO:   04:23:40 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 3e0195c3-2a0c-493f-ba94-832693d7a6b3 in 0.45 seconds
INFO:   04:23:40 - uvicorn.access - 172.17.0.1:58286 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:23:40 - uvicorn.access - 172.17.0.1:58286 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:23:40 - uvicorn.access - 172.17.0.1:58286 - "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:   04:23:25 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:25 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:23:25 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:23:25 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:23:25 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/readyz$,/livez$,/health$,/ready$,/healthz$,/metrics$)
INFO:   04:23:25 - uvicorn.error - Started server process [1]
INFO:   04:23:25 - uvicorn.error - Waiting for application startup.
INFO:   04:23:28 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:28 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:23:28 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:28 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:23:29 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:29 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 118619a1d3ed3b201b814dac997d6742
INFO:   04:23:29 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:23:29 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:23:29 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:23:29.175913768 [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]+
INFO:   04:23:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:23:29,418 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:29,420 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:23:29,479 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:29,479 [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] 2026-10-05 04:23:29,580 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:29,580 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:23:29 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:29 - 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, 9843.08it/s]
INFO:   04:23:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:31 - uvicorn.error - Application startup complete.
INFO:   04:23:31 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:23:31 - docling_serve.app - Health check requested
INFO:   04:23:31 - uvicorn.access - 172.17.0.1:37930 - "GET /health HTTP/1.1" 200
INFO:   04:23:31 - docling_serve.app - Health check requested
INFO:   04:23:31 - uvicorn.access - 172.17.0.1:37932 - "GET /health HTTP/1.1" 200
INFO:   04:23:31 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:23:31 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:23:31 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:23:31 - docling_serve.app - [TENANT_ID] Task e53b1e80-00eb-4074-97f4-e1c2a04570fa created with tenant_id='default'
INFO:   04:23:31 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task e53b1e80-00eb-4074-97f4-e1c2a04570fa
INFO:   04:23:31 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:31 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 83f6e7119a1499def48d21cd30b03869
INFO:   04:23:31 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:23:31 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:23:31,667 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:31,668 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:23:31,786 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:31,786 [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] 2026-10-05 04:23:31,836 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:31,836 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:23:31 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:31 - 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, 5268.18it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5288.38it/s]
INFO:   04:23:32 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:33 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:23:33 - docling.document_converter - Going to convert document batch...
INFO:   04:23:33 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:23:33 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:23:33 - docling.document_converter - Finished converting document file in 0.60 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:    04:23:33 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:23:33 - docling_jobkit.convert.results - Processed 1 docs in 0.67 seconds.
INFO:   04:23:33 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job e53b1e80-00eb-4074-97f4-e1c2a04570fa in 0.67 seconds
INFO:   04:23:33 - uvicorn.access - 172.17.0.1:37932 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:23:33 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:23:33 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:23:33 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:23:33 - docling_serve.app - [TENANT_ID] Task 6deac06a-59d6-414b-b86a-4c3bd15ade34 created with tenant_id='default'
INFO:   04:23:33 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 6deac06a-59d6-414b-b86a-4c3bd15ade34
INFO:   04:23:33 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:33 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 83f6e7119a1499def48d21cd30b03869
INFO:   04:23:33 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:23:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:23:33,657 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:33,657 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:23:33,713 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:33,713 [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] 2026-10-05 04:23:33,816 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:33,817 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:23:33 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:33 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  64%|██████▍   | 493/770 [00:00<00:00, 4928.86it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4804.45it/s]
INFO:   04:23:34 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:34 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:23:35 - docling.document_converter - Going to convert document batch...
INFO:   04:23:35 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:23:35 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:23:35 - docling.document_converter - Finished converting document file in 0.17 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:   04:23:35 - docling_jobkit.convert.results - Processed 1 docs in 0.18 seconds.
INFO:   04:23:35 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 6deac06a-59d6-414b-b86a-4c3bd15ade34 in 0.18 seconds
INFO:   04:23:35 - uvicorn.access - 172.17.0.1:37932 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:23:35 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:23:35 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:23:35 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:23:35 - docling_serve.app - [TENANT_ID] Task 0ba082f1-9a62-4b8f-a8c5-c919fabb451a created with tenant_id='default'
INFO:   04:23:35 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 0ba082f1-9a62-4b8f-a8c5-c919fabb451a
INFO:   04:23:35 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:23:35 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 528ff53b5cc66cd5f7f61ef2cd72a8fc
INFO:   04:23:35 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:23:35 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:23:35,679 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:35,680 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/det/PP-OCRv6_det_small.onnx
[INFO] 2026-10-05 04:23:35,726 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:35,726 [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] 2026-10-05 04:23:35,783 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:23:35,783 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv6/rec/PP-OCRv6_rec_small.onnx
INFO:   04:23:35 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:23:35 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  69%|██████▉   | 534/770 [00:00<00:00, 5332.14it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5359.06it/s]
INFO:   04:23:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:23:36 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:23:36 - docling.document_converter - Going to convert document batch...
INFO:   04:23:36 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:23:36 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:23:36 - docling.document_converter - Finished converting document file in 0.42 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:   04:23:36 - docling_jobkit.convert.results - Processed 1 docs in 0.45 seconds.
INFO:   04:23:36 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 0ba082f1-9a62-4b8f-a8c5-c919fabb451a in 0.45 seconds
INFO:   04:23:37 - uvicorn.access - 172.17.0.1:37932 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:23:37 - uvicorn.access - 172.17.0.1:37932 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:23:37 - uvicorn.access - 172.17.0.1:37932 - "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:   04:20:35 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:20:35 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'nemotron-ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:20:35 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:20:35 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:20:35 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/readyz$,/livez$,/healthz$,/ready$,/metrics$,/health$)
INFO:   04:20:36 - uvicorn.error - Started server process [1]
INFO:   04:20:36 - uvicorn.error - Waiting for application startup.
INFO:   04:20:39 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:20:39 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:20:39 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:20:39 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:20:39 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:20:40 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c4d6547d10e26a777a530d63ca01a1a1
INFO:   04:20:40 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:20:40 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
INFO:   04:20:40 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
2026-10-05 04:20:40.103693624 [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]+
INFO:   04:20:40 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:20:40,339 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:40,341 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:20:40,434 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:40,434 [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] 2026-10-05 04:20:40,480 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:40,480 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:20:40 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:20:40 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  70%|██████▉   | 536/770 [00:00<00:00, 5357.71it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5421.61it/s]
INFO:   04:20:41 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:20:42 - uvicorn.error - Application startup complete.
INFO:   04:20:42 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:20:43 - docling_serve.app - Health check requested
INFO:   04:20:43 - uvicorn.access - 172.17.0.1:41254 - "GET /health HTTP/1.1" 200
INFO:   04:20:43 - docling_serve.app - Health check requested
INFO:   04:20:43 - uvicorn.access - 172.17.0.1:41262 - "GET /health HTTP/1.1" 200
INFO:   04:20:43 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:20:43 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:20:43 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:20:43 - docling_serve.app - [TENANT_ID] Task 73015cac-d5ce-4b5e-9908-975fe3a156f1 created with tenant_id='default'
INFO:   04:20:43 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 73015cac-d5ce-4b5e-9908-975fe3a156f1
INFO:   04:20:43 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:20:43 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 4ef9920c0fc80ed745870d6025fc17ac
INFO:   04:20:43 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:20:43 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:20:43,166 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:43,166 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:20:43,267 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:43,267 [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] 2026-10-05 04:20:43,317 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:43,317 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:20:43 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:20:43 - 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, 9234.51it/s]
INFO:   04:20:43 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:20:44 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:20:44 - docling.document_converter - Going to convert document batch...
INFO:   04:20:44 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:20:44 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:20:44 - docling.document_converter - Finished converting document file in 0.41 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:    04:20:44 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:20:44 - docling_jobkit.convert.results - Processed 1 docs in 0.48 seconds.
INFO:   04:20:44 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 73015cac-d5ce-4b5e-9908-975fe3a156f1 in 0.48 seconds
INFO:   04:20:45 - uvicorn.access - 172.17.0.1:41262 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:20:45 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:20:45 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:20:45 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:20:45 - docling_serve.app - [TENANT_ID] Task ac601233-48b4-43f4-8165-6a07325e85d7 created with tenant_id='default'
INFO:   04:20:45 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task ac601233-48b4-43f4-8165-6a07325e85d7
INFO:   04:20:45 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:20:45 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 4ef9920c0fc80ed745870d6025fc17ac
INFO:   04:20:45 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:20:45 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:20:45,182 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:45,182 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:20:45,291 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:45,292 [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] 2026-10-05 04:20:45,340 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:45,340 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:20:45 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:20:45 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  67%|██████▋   | 518/770 [00:00<00:00, 5173.59it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5223.26it/s]
INFO:   04:20:45 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:20:46 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:20:46 - docling.document_converter - Going to convert document batch...
INFO:   04:20:46 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:20:46 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:20:46 - docling.document_converter - Finished converting document file in 0.68 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:   04:20:46 - docling_jobkit.convert.results - Processed 1 docs in 0.68 seconds.
INFO:   04:20:46 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job ac601233-48b4-43f4-8165-6a07325e85d7 in 0.68 seconds
INFO:   04:20:47 - uvicorn.access - 172.17.0.1:41262 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:20:47 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:20:47 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:20:47 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:20:47 - docling_serve.app - [TENANT_ID] Task 5521760c-d661-474b-a2da-cc2fd58fa66c created with tenant_id='default'
INFO:   04:20:47 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 5521760c-d661-474b-a2da-cc2fd58fa66c
INFO:   04:20:47 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:20:47 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 63e8839d1103b03d45382915dba33c89
INFO:   04:20:47 - docling.models.stages.ocr.auto_ocr_model - Nemotron cannot be used because it is not installed.
INFO:   04:20:47 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:20:47,206 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:47,206 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:20:47,297 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:47,297 [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] 2026-10-05 04:20:47,346 [RapidOCR] base.py:23: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:47,346 [RapidOCR] main.py:63: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:20:47 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:20:47 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  58%|█████▊    | 446/770 [00:00<00:00, 4457.43it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4433.62it/s]
INFO:   04:20:47 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:20:48 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:20:48 - docling.document_converter - Going to convert document batch...
INFO:   04:20:48 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:20:48 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:20:48 - docling.document_converter - Finished converting document file in 0.26 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:   04:20:48 - docling_jobkit.convert.results - Processed 1 docs in 0.30 seconds.
INFO:   04:20:48 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 5521760c-d661-474b-a2da-cc2fd58fa66c in 0.30 seconds
INFO:   04:20:49 - uvicorn.access - 172.17.0.1:41262 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:20:49 - uvicorn.access - 172.17.0.1:41262 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:20:49 - uvicorn.access - 172.17.0.1:41262 - "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:   04:20:32 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:20:32 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:20:32 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:20:32 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:20:32 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/healthz$,/health$,/ready$,/metrics$,/readyz$,/livez$)
INFO:   04:20:32 - uvicorn.error - Started server process [1]
INFO:   04:20:32 - uvicorn.error - Waiting for application startup.
INFO:   04:20:35 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:20:35 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:20:35 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:20:35 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:20:35 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:20:35 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash f446dcc5a1a7c6aff458e30cab474a9e
INFO:   04:20:35 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:20:35 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-10-05 04:20:35.909319562 [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]+
INFO:   04:20:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:20:36,253 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:36,255 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:20:36,394 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:36,394 [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] 2026-10-05 04:20:36,447 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:36,448 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:20:36 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:20:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▌   | 504/770 [00:00<00:00, 5007.55it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4981.88it/s]
INFO:   04:20:37 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:20:38 - uvicorn.error - Application startup complete.
INFO:   04:20:38 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:20:39 - docling_serve.app - Health check requested
INFO:   04:20:39 - uvicorn.access - 172.17.0.1:55628 - "GET /health HTTP/1.1" 200
INFO:   04:20:39 - docling_serve.app - Health check requested
INFO:   04:20:39 - uvicorn.access - 172.17.0.1:55642 - "GET /health HTTP/1.1" 200
INFO:   04:20:39 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:20:39 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:20:39 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:20:39 - docling_serve.app - [TENANT_ID] Task f5c2eeab-4243-442a-b6e7-83840d153353 created with tenant_id='default'
INFO:   04:20:39 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task f5c2eeab-4243-442a-b6e7-83840d153353
INFO:   04:20:39 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:20:39 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c639c48e78aacd42b6243acf42ab1c3a
INFO:   04:20:39 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:20:39,152 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:39,152 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:20:39,247 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:39,248 [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] 2026-10-05 04:20:39,295 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:39,295 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:20:39 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:20:39 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  71%|███████   | 548/770 [00:00<00:00, 5471.15it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5459.42it/s]
INFO:   04:20:39 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:20:40 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:20:40 - docling.document_converter - Going to convert document batch...
INFO:   04:20:40 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:20:40 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:20:40 - docling.document_converter - Finished converting document file in 0.61 sec.
WARNING:    04:20:40 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:20:40 - docling_jobkit.convert.results - Processed 1 docs in 0.69 seconds.
INFO:   04:20:40 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job f5c2eeab-4243-442a-b6e7-83840d153353 in 0.69 seconds
INFO:   04:20:41 - uvicorn.access - 172.17.0.1:55642 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:20:41 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:20:41 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:20:41 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:20:41 - docling_serve.app - [TENANT_ID] Task 610fbb1d-7077-47b0-a70c-8a8ddd7cfef0 created with tenant_id='default'
INFO:   04:20:41 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 610fbb1d-7077-47b0-a70c-8a8ddd7cfef0
INFO:   04:20:41 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:20:41 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c639c48e78aacd42b6243acf42ab1c3a
INFO:   04:20:41 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:20:41,611 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:41,611 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:20:41,722 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:41,722 [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] 2026-10-05 04:20:41,869 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:41,873 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:20:42 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:20:42 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  69%|██████▉   | 533/770 [00:00<00:00, 5329.10it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5374.81it/s]
INFO:   04:20:42 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:20:43 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:20:43 - docling.document_converter - Going to convert document batch...
INFO:   04:20:43 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:20:43 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:20:43 - docling.document_converter - Finished converting document file in 0.71 sec.
INFO:   04:20:43 - docling_jobkit.convert.results - Processed 1 docs in 0.71 seconds.
INFO:   04:20:43 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 610fbb1d-7077-47b0-a70c-8a8ddd7cfef0 in 0.71 seconds
INFO:   04:20:43 - uvicorn.access - 172.17.0.1:55642 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:20:43 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:20:43 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:20:43 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:20:43 - docling_serve.app - [TENANT_ID] Task 440836b5-f7c6-4189-9501-b3091bc9f0d2 created with tenant_id='default'
INFO:   04:20:43 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 440836b5-f7c6-4189-9501-b3091bc9f0d2
INFO:   04:20:43 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:20:43 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash e50b6140ec7ee9be0d0fea6ba20c9795
INFO:   04:20:43 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:20:43,591 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:43,591 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:20:43,697 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:43,697 [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] 2026-10-05 04:20:43,777 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:20:43,777 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:20:43 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:20:43 - 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, 5050.31it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5122.20it/s]
INFO:   04:20:44 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:20:44 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:20:44 - docling.document_converter - Going to convert document batch...
INFO:   04:20:44 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:20:44 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:20:44 - docling.document_converter - Finished converting document file in 0.37 sec.
INFO:   04:20:45 - docling_jobkit.convert.results - Processed 1 docs in 0.40 seconds.
INFO:   04:20:45 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 440836b5-f7c6-4189-9501-b3091bc9f0d2 in 0.40 seconds
INFO:   04:20:45 - uvicorn.access - 172.17.0.1:55642 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:20:45 - uvicorn.access - 172.17.0.1:55642 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:20:45 - uvicorn.access - 172.17.0.1:55642 - "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:   04:17:52 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:17:52 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:17:52 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:17:52 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:17:52 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/ready$,/healthz$,/livez$,/readyz$,/health$)
INFO:   04:17:53 - uvicorn.error - Started server process [1]
INFO:   04:17:53 - uvicorn.error - Waiting for application startup.
INFO:   04:17:56 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:17:56 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:17:56 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:17:56 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:17:56 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:17:56 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 1743cd6dac10356c82ba7dfffff4ed02
INFO:   04:17:56 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:17:56 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-10-05 04:17:56.981011927 [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]+
INFO:   04:17:57 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:17:57,343 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:57,345 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:17:57,524 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:57,524 [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] 2026-10-05 04:17:57,598 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:57,598 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:17:57 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:17:57 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  68%|██████▊   | 520/770 [00:00<00:00, 5189.40it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5293.89it/s]
INFO:   04:17:58 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:17:58 - uvicorn.error - Application startup complete.
INFO:   04:17:58 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:17:59 - docling_serve.app - Health check requested
INFO:   04:17:59 - uvicorn.access - 172.17.0.1:37266 - "GET /health HTTP/1.1" 200
INFO:   04:17:59 - docling_serve.app - Health check requested
INFO:   04:17:59 - uvicorn.access - 172.17.0.1:37274 - "GET /health HTTP/1.1" 200
INFO:   04:17:59 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:17:59 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:17:59 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:17:59 - docling_serve.app - [TENANT_ID] Task c8a8375c-a7c6-47ed-a10a-611d8badf9e7 created with tenant_id='default'
INFO:   04:17:59 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task c8a8375c-a7c6-47ed-a10a-611d8badf9e7
INFO:   04:17:59 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:17:59 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 447bb8cb2d04f05cb199adb5a03148da
INFO:   04:17:59 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:17:59,218 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:59,218 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:17:59,345 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:59,345 [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] 2026-10-05 04:17:59,393 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:59,393 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:17:59 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:17:59 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  64%|██████▍   | 491/770 [00:00<00:00, 4906.69it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5593.79it/s]
INFO:   04:17:59 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:18:00 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:18:00 - docling.document_converter - Going to convert document batch...
INFO:   04:18:00 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:18:00 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:18:00 - docling.document_converter - Finished converting document file in 0.46 sec.
WARNING:    04:18:00 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:18:00 - docling_jobkit.convert.results - Processed 1 docs in 0.56 seconds.
INFO:   04:18:00 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job c8a8375c-a7c6-47ed-a10a-611d8badf9e7 in 0.56 seconds
INFO:   04:18:01 - uvicorn.access - 172.17.0.1:37274 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:18:01 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:18:01 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:18:01 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:18:01 - docling_serve.app - [TENANT_ID] Task 42586483-df33-4ea2-8735-6b836b153da1 created with tenant_id='default'
INFO:   04:18:01 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 42586483-df33-4ea2-8735-6b836b153da1
INFO:   04:18:01 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:18:01 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 447bb8cb2d04f05cb199adb5a03148da
INFO:   04:18:01 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:18:01,545 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:18:01,545 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:18:01,691 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:18:01,692 [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] 2026-10-05 04:18:01,745 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:18:01,745 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:18:01 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:18:01 - 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, 4584.64it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4545.65it/s]
INFO:   04:18:02 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:18:03 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:18:03 - docling.document_converter - Going to convert document batch...
INFO:   04:18:03 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:18:03 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:18:03 - docling.document_converter - Finished converting document file in 0.50 sec.
INFO:   04:18:03 - docling_jobkit.convert.results - Processed 1 docs in 0.51 seconds.
INFO:   04:18:03 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 42586483-df33-4ea2-8735-6b836b153da1 in 0.51 seconds
INFO:   04:18:03 - uvicorn.access - 172.17.0.1:37274 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:18:03 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:18:03 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:18:03 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:18:03 - docling_serve.app - [TENANT_ID] Task 441bacf6-89d9-4d70-b76a-ec7ca134f3f5 created with tenant_id='default'
INFO:   04:18:03 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 441bacf6-89d9-4d70-b76a-ec7ca134f3f5
INFO:   04:18:03 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:18:03 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 6102b5a2a09ce02f29b31fc27cc3d04f
INFO:   04:18:03 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:18:03,560 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:18:03,560 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:18:03,685 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:18:03,686 [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] 2026-10-05 04:18:03,734 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:18:03,734 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:18:03 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:18:03 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  62%|██████▏   | 475/770 [00:00<00:00, 4748.32it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4657.37it/s]
INFO:   04:18:04 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:18:04 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:18:04 - docling.document_converter - Going to convert document batch...
INFO:   04:18:04 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:18:04 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:18:04 - docling.document_converter - Finished converting document file in 0.29 sec.
INFO:   04:18:04 - docling_jobkit.convert.results - Processed 1 docs in 0.33 seconds.
INFO:   04:18:04 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 441bacf6-89d9-4d70-b76a-ec7ca134f3f5 in 0.33 seconds
INFO:   04:18:05 - uvicorn.access - 172.17.0.1:37274 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:18:05 - uvicorn.access - 172.17.0.1:37274 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:18:05 - uvicorn.access - 172.17.0.1:37274 - "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:   04:17:46 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:17:46 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:17:46 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:17:46 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:17:46 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/healthz$,/ready$,/livez$,/readyz$,/metrics$,/health$)
INFO:   04:17:46 - uvicorn.error - Started server process [1]
INFO:   04:17:46 - uvicorn.error - Waiting for application startup.
INFO:   04:17:50 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:17:50 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:17:50 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:17:50 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:17:50 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:17:50 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 1743cd6dac10356c82ba7dfffff4ed02
INFO:   04:17:50 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:17:50 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-10-05 04:17:50.960497228 [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]+
INFO:   04:17:51 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:17:51,318 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:51,325 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:17:51,453 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:51,453 [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] 2026-10-05 04:17:51,534 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:51,535 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:17:51 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:17:51 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  68%|██████▊   | 521/770 [00:00<00:00, 5205.56it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5112.17it/s]
INFO:   04:17:52 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:17:52 - uvicorn.error - Application startup complete.
INFO:   04:17:52 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:17:53 - docling_serve.app - Health check requested
INFO:   04:17:53 - uvicorn.access - 172.17.0.1:48516 - "GET /health HTTP/1.1" 200
INFO:   04:17:53 - docling_serve.app - Health check requested
INFO:   04:17:53 - uvicorn.access - 172.17.0.1:48528 - "GET /health HTTP/1.1" 200
INFO:   04:17:53 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:17:53 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:17:53 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:17:53 - docling_serve.app - [TENANT_ID] Task bfcccacd-ac13-4738-a397-2863393ce438 created with tenant_id='default'
INFO:   04:17:53 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task bfcccacd-ac13-4738-a397-2863393ce438
INFO:   04:17:53 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:17:53 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 447bb8cb2d04f05cb199adb5a03148da
INFO:   04:17:53 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:17:53,219 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:53,220 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:17:53,314 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:53,314 [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] 2026-10-05 04:17:53,361 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:53,361 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:17:53 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:17:53 - 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, 5018.01it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5023.41it/s]
INFO:   04:17:53 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:17:54 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:17:54 - docling.document_converter - Going to convert document batch...
INFO:   04:17:54 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:17:54 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:17:54 - docling.document_converter - Finished converting document file in 0.66 sec.
WARNING:    04:17:54 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:17:54 - docling_jobkit.convert.results - Processed 1 docs in 0.73 seconds.
INFO:   04:17:54 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job bfcccacd-ac13-4738-a397-2863393ce438 in 0.73 seconds
INFO:   04:17:55 - uvicorn.access - 172.17.0.1:48528 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:17:55 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:17:55 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:17:55 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:17:55 - docling_serve.app - [TENANT_ID] Task 4b428dc0-1f88-4c79-b4a3-415c8a664238 created with tenant_id='default'
INFO:   04:17:55 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 4b428dc0-1f88-4c79-b4a3-415c8a664238
INFO:   04:17:55 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:17:55 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 447bb8cb2d04f05cb199adb5a03148da
INFO:   04:17:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:17:55,275 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:55,275 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:17:55,368 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:55,368 [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] 2026-10-05 04:17:55,415 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:55,415 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:17:55 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:17:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  68%|██████▊   | 524/770 [00:00<00:00, 5239.26it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5121.88it/s]
INFO:   04:17:55 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:17:56 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:17:56 - docling.document_converter - Going to convert document batch...
INFO:   04:17:56 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:17:56 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:17:56 - docling.document_converter - Finished converting document file in 0.35 sec.
INFO:   04:17:56 - docling_jobkit.convert.results - Processed 1 docs in 0.36 seconds.
INFO:   04:17:56 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 4b428dc0-1f88-4c79-b4a3-415c8a664238 in 0.36 seconds
INFO:   04:17:57 - uvicorn.access - 172.17.0.1:48528 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:17:57 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:17:57 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:17:57 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:17:57 - docling_serve.app - [TENANT_ID] Task 2ab516a5-84a6-451c-987f-6f552e87d720 created with tenant_id='default'
INFO:   04:17:57 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 2ab516a5-84a6-451c-987f-6f552e87d720
INFO:   04:17:57 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:17:57 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 6102b5a2a09ce02f29b31fc27cc3d04f
INFO:   04:17:57 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:17:57,254 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:57,254 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:17:57,346 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:57,346 [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] 2026-10-05 04:17:57,465 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:17:57,465 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:17:57 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:17:57 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  65%|██████▍   | 497/770 [00:00<00:00, 4968.37it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4975.38it/s]
INFO:   04:17:58 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:17:58 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:17:58 - docling.document_converter - Going to convert document batch...
INFO:   04:17:58 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:17:58 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:17:58 - docling.document_converter - Finished converting document file in 0.40 sec.
INFO:   04:17:58 - docling_jobkit.convert.results - Processed 1 docs in 0.45 seconds.
INFO:   04:17:58 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 2ab516a5-84a6-451c-987f-6f552e87d720 in 0.45 seconds
INFO:   04:17:59 - uvicorn.access - 172.17.0.1:48528 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:17:59 - uvicorn.access - 172.17.0.1:48528 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:17:59 - uvicorn.access - 172.17.0.1:48528 - "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:   04:15:31 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:15:31 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:15:31 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:15:31 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:15:31 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/ready$,/readyz$,/metrics$,/healthz$,/livez$,/health$)
INFO:   04:15:31 - uvicorn.error - Started server process [1]
INFO:   04:15:31 - uvicorn.error - Waiting for application startup.
INFO:   04:15:36 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:15:36 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:15:36 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:15:36 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:15:36 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:15:36 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 1743cd6dac10356c82ba7dfffff4ed02
INFO:   04:15:36 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:15:36 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-10-05 04:15:36.272615332 [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]+
INFO:   04:15:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:15:36,643 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:36,644 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:15:36,740 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:36,740 [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] 2026-10-05 04:15:36,786 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:36,786 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:15:36 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:15:36 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  59%|█████▊    | 451/770 [00:00<00:00, 4474.42it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4445.56it/s]
INFO:   04:15:37 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:15:37 - uvicorn.error - Application startup complete.
INFO:   04:15:37 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:15:38 - docling_serve.app - Health check requested
INFO:   04:15:38 - uvicorn.access - 172.17.0.1:38758 - "GET /health HTTP/1.1" 200
INFO:   04:15:38 - docling_serve.app - Health check requested
INFO:   04:15:38 - uvicorn.access - 172.17.0.1:38768 - "GET /health HTTP/1.1" 200
INFO:   04:15:38 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:15:38 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:15:38 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:15:38 - docling_serve.app - [TENANT_ID] Task 9976e18a-d8b6-4dad-b96a-8cb3531f2fe2 created with tenant_id='default'
INFO:   04:15:38 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 9976e18a-d8b6-4dad-b96a-8cb3531f2fe2
INFO:   04:15:38 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:15:38 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a1cb01c48360ff351eb870dc682aac59
INFO:   04:15:38 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:15:38,575 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:38,576 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:15:38,670 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:38,670 [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] 2026-10-05 04:15:38,751 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:38,751 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:15:38 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:15:38 - 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, 4615.11it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4500.11it/s]
INFO:   04:15:39 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:15:39 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:15:39 - docling.document_converter - Going to convert document batch...
INFO:   04:15:39 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:15:39 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:15:39 - docling.document_converter - Finished converting document file in 0.33 sec.
WARNING:    04:15:40 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:15:40 - docling_jobkit.convert.results - Processed 1 docs in 0.72 seconds.
INFO:   04:15:40 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 9976e18a-d8b6-4dad-b96a-8cb3531f2fe2 in 0.72 seconds
INFO:   04:15:40 - uvicorn.access - 172.17.0.1:38768 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:15:40 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:15:40 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:15:40 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:15:40 - docling_serve.app - [TENANT_ID] Task 4e81f40f-0f40-4043-a066-062150944db3 created with tenant_id='default'
INFO:   04:15:40 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task 4e81f40f-0f40-4043-a066-062150944db3
INFO:   04:15:40 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:15:40 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash a1cb01c48360ff351eb870dc682aac59
INFO:   04:15:40 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:15:40,588 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:40,588 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:15:40,705 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:40,705 [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] 2026-10-05 04:15:40,752 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:40,752 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:15:40 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:15:40 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  60%|██████    | 463/770 [00:00<00:00, 4616.48it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4547.84it/s]
INFO:   04:15:41 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:15:41 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:15:42 - docling.document_converter - Going to convert document batch...
INFO:   04:15:42 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:15:42 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:15:42 - docling.document_converter - Finished converting document file in 0.39 sec.
INFO:   04:15:42 - docling_jobkit.convert.results - Processed 1 docs in 0.40 seconds.
INFO:   04:15:42 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job 4e81f40f-0f40-4043-a066-062150944db3 in 0.40 seconds
INFO:   04:15:42 - uvicorn.access - 172.17.0.1:38768 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:15:42 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:15:42 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:15:42 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:15:42 - docling_serve.app - [TENANT_ID] Task e698acd5-e95f-4204-b8ff-555ff8f51bd4 created with tenant_id='default'
INFO:   04:15:42 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task e698acd5-e95f-4204-b8ff-555ff8f51bd4
INFO:   04:15:42 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:15:42 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash fb1751b09de6bdbe35ede48676ca8f82
INFO:   04:15:42 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:15:42,603 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:42,603 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:15:42,703 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:42,703 [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] 2026-10-05 04:15:42,758 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:42,758 [RapidOCR] main.py:65: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:15:42 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:15:42 - 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, 4750.84it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5109.39it/s]
INFO:   04:15:43 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:15:43 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:15:43 - docling.document_converter - Going to convert document batch...
INFO:   04:15:43 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:15:43 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:15:44 - docling.document_converter - Finished converting document file in 0.40 sec.
INFO:   04:15:44 - docling_jobkit.convert.results - Processed 1 docs in 0.46 seconds.
INFO:   04:15:44 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job e698acd5-e95f-4204-b8ff-555ff8f51bd4 in 0.46 seconds
INFO:   04:15:44 - uvicorn.access - 172.17.0.1:38768 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:15:44 - uvicorn.access - 172.17.0.1:38768 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:15:44 - uvicorn.access - 172.17.0.1:38768 - "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:   04:15:16 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:15:16 - docling.models.factories - Registered ocr engines: ['auto', 'easyocr', 'kserve_v2_ocr', 'ocrmac', 'rapidocr', 'tesserocr', 'tesseract']
INFO:   04:15:16 - docling_serve.otel_instrumentation - Setting up OpenTelemetry metrics
INFO:   04:15:16 - docling_serve.otel_instrumentation - Enabling Prometheus metrics export
INFO:   04:15:16 - docling_serve.otel_instrumentation - Instrumenting FastAPI with OpenTelemetry (excluded_urls=/metrics$,/health$,/readyz$,/ready$,/livez$,/healthz$)
INFO:   04:15:16 - uvicorn.error - Started server process [1]
INFO:   04:15:16 - uvicorn.error - Waiting for application startup.
INFO:   04:15:23 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:15:23 - docling.models.factories - Registered table structure engines: ['docling_tableformer', 'docling_tableformer_v2', 'granite_vision_table']
INFO:   04:15:23 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:15:23 - docling.models.factories - Registered layout engines: ['layout_object_detection', 'docling_layout_default', 'docling_experimental_table_crops_layout']
INFO:   04:15:23 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:15:23 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash bcb2e23b24bd94a42b5c00d662a7c1cf
INFO:   04:15:23 - docling.models.factories.base_factory - Loading plugin 'docling_defaults'
INFO:   04:15:23 - docling.models.factories - Registered picture descriptions: ['picture_description_vlm_engine', 'vlm', 'api']
2026-10-05 04:15:23.195258737 [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]+
INFO:   04:15:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:15:23,603 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:23,605 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:15:23,741 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:23,741 [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] 2026-10-05 04:15:23,793 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:23,793 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:15:23 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:15:23 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  61%|██████    | 471/770 [00:00<00:00, 4688.19it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4567.68it/s]
INFO:   04:15:24 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:15:25 - uvicorn.error - Application startup complete.
INFO:   04:15:25 - uvicorn.error - Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:   04:15:25 - docling_serve.app - Health check requested
INFO:   04:15:25 - uvicorn.access - 172.17.0.1:51624 - "GET /health HTTP/1.1" 200
INFO:   04:15:25 - docling_serve.app - Health check requested
INFO:   04:15:25 - uvicorn.access - 172.17.0.1:51632 - "GET /health HTTP/1.1" 200
INFO:   04:15:25 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:15:25 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:15:25 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:15:25 - docling_serve.app - [TENANT_ID] Task 331fff8d-f77a-4b88-8211-684b0e70f07b created with tenant_id='default'
INFO:   04:15:25 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task 331fff8d-f77a-4b88-8211-684b0e70f07b
INFO:   04:15:25 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:15:25 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c835e945376007df084ecda847e91f5b
INFO:   04:15:25 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:15:25,556 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:25,557 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:15:25,666 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:25,666 [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] 2026-10-05 04:15:25,715 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:25,715 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:15:25 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:15:25 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  56%|█████▌    | 428/770 [00:00<00:00, 4260.23it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4374.70it/s]
INFO:   04:15:26 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:15:26 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:15:27 - docling.document_converter - Going to convert document batch...
INFO:   04:15:27 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:15:27 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:15:27 - docling.document_converter - Finished converting document file in 0.42 sec.
WARNING:    04:15:27 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
INFO:   04:15:27 - docling_jobkit.convert.results - Processed 1 docs in 0.51 seconds.
INFO:   04:15:27 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job 331fff8d-f77a-4b88-8211-684b0e70f07b in 0.51 seconds
INFO:   04:15:27 - uvicorn.access - 172.17.0.1:51632 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:15:27 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:15:27 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:15:27 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:15:27 - docling_serve.app - [TENANT_ID] Task e08efe9f-5d79-416e-80d3-03813aed73de created with tenant_id='default'
INFO:   04:15:27 - docling_jobkit.orchestrators.local.worker - Worker 1 processing task e08efe9f-5d79-416e-80d3-03813aed73de
INFO:   04:15:27 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:15:27 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash c835e945376007df084ecda847e91f5b
INFO:   04:15:27 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:15:27,898 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:27,898 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:15:28,014 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:28,014 [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] 2026-10-05 04:15:28,090 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:28,091 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:15:28 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:15:28 - docling.utils.accelerator_utils - Accelerator device: 'cpu'

Loading weights:   0%|          | 0/770 [00:00<?, ?it/s]
Loading weights:  64%|██████▎   | 490/770 [00:00<00:00, 4897.41it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4921.36it/s]
INFO:   04:15:28 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:15:29 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:15:29 - docling.document_converter - Going to convert document batch...
INFO:   04:15:29 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:15:29 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:15:29 - docling.document_converter - Finished converting document file in 0.39 sec.
INFO:   04:15:29 - docling_jobkit.convert.results - Processed 1 docs in 0.41 seconds.
INFO:   04:15:29 - docling_jobkit.orchestrators.local.worker - Worker 1 completed job e08efe9f-5d79-416e-80d3-03813aed73de in 0.41 seconds
INFO:   04:15:29 - uvicorn.access - 172.17.0.1:51632 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:15:29 - docling_serve.app - [TENANT_ID] Extracted tenant_id from header: 'default' (header_value: 'None')
INFO:   04:15:29 - docling_serve.app - [TENANT_ID] process_url endpoint received tenant_id='default'
INFO:   04:15:29 - docling_serve.app - [TENANT_ID] Preparing to enqueue with tenant_id='default' in metadata
INFO:   04:15:29 - docling_serve.app - [TENANT_ID] Task ead2b864-8eab-4c10-9e15-9cd3894b21af created with tenant_id='default'
INFO:   04:15:29 - docling_jobkit.orchestrators.local.worker - Worker 0 processing task ead2b864-8eab-4c10-9e15-9cd3894b21af
INFO:   04:15:29 - docling_jobkit.convert.manager - artifacts_path is set to a valid directory. No model weights will be downloaded at runtime.
INFO:   04:15:29 - docling.document_converter - Initializing pipeline for StandardPdfPipeline with options hash 3eabf304acb4a56e9ef52b02ef69bccb
INFO:   04:15:29 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
[INFO] 2026-10-05 04:15:29,904 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:29,905 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:15:30,022 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:30,022 [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] 2026-10-05 04:15:30,096 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:15:30,096 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/rec/ch_PP-OCRv4_rec_mobile.onnx
INFO:   04:15:30 - docling.models.stages.ocr.auto_ocr_model - Auto OCR model selected rapidocr with onnxruntime.
INFO:   04:15:30 - 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, 5009.73it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4988.51it/s]
INFO:   04:15:30 - docling.utils.accelerator_utils - Accelerator device: 'cpu'
INFO:   04:15:31 - docling.datamodel.document - detected formats: [<InputFormat.HTML: 'html'>]
INFO:   04:15:31 - docling.document_converter - Going to convert document batch...
INFO:   04:15:31 - docling.document_converter - Initializing pipeline for SimplePipeline with options hash 7d306d2d021deac65a97d1a5f925362a
INFO:   04:15:31 - docling.pipeline.base_pipeline - Processing document file
INFO:   04:15:31 - docling.document_converter - Finished converting document file in 0.36 sec.
INFO:   04:15:31 - docling_jobkit.convert.results - Processed 1 docs in 0.41 seconds.
INFO:   04:15:31 - docling_jobkit.orchestrators.local.worker - Worker 0 completed job ead2b864-8eab-4c10-9e15-9cd3894b21af in 0.41 seconds
INFO:   04:15:31 - uvicorn.access - 172.17.0.1:51632 - "POST /v1/convert/source HTTP/1.1" 200
INFO:   04:15:32 - uvicorn.access - 172.17.0.1:51632 - "GET /v1/clear/converters HTTP/1.1" 200
INFO:   04:15:32 - uvicorn.access - 172.17.0.1:51632 - "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:
2026-10-05 04:12:48.742531212 [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]+
[INFO] 2026-10-05 04:12:49,116 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:49,118 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:12:49,238 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:49,239 [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] 2026-10-05 04:12:49,306 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:49,306 [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:  75%|███████▌  | 579/770 [00:00<00:00, 5772.21it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5360.80it/s]
[INFO] 2026-10-05 04:12:51,208 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:51,209 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:12:51,332 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:51,332 [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] 2026-10-05 04:12:51,380 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:51,380 [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%|█████▉    | 455/770 [00:00<00:00, 4526.27it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4520.91it/s]
WARNING:    04:12:52 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
WARNING:    04:12:52 - docling_core.types.doc.document - Parameter `strict_text` has been deprecated and will be ignored.
[INFO] 2026-10-05 04:12:53,527 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:53,528 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:12:53,642 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:53,642 [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] 2026-10-05 04:12:53,688 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:53,688 [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%|██████▊   | 525/770 [00:00<00:00, 5249.42it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5211.52it/s]
[INFO] 2026-10-05 04:12:55,263 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:55,264 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:12:55,431 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:55,431 [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] 2026-10-05 04:12:55,536 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:55,537 [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%|██████▏   | 476/770 [00:00<00:00, 4282.11it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5597.64it/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.
2026-10-05 04:12:33.208090662 [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]+
[INFO] 2026-10-05 04:12:33,722 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:33,725 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:12:33,863 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:33,864 [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] 2026-10-05 04:12:33,912 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:33,913 [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:  66%|██████▌   | 510/770 [00:00<00:00, 5096.36it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5171.76it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:39288 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39290 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:12:36,449 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:36,449 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:12:36,547 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:36,547 [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] 2026-10-05 04:12:36,596 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:36,596 [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:  69%|██████▉   | 535/770 [00:00<00:00, 5344.78it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5382.90it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39290 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:12:40,624 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:40,624 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:12:40,745 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:40,745 [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] 2026-10-05 04:12:40,792 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:40,792 [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%|██████▌   | 503/770 [00:00<00:00, 5026.88it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5004.44it/s]
INFO:     172.17.0.1:39290 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:12:42,637 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:42,638 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:12:42,759 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:42,759 [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] 2026-10-05 04:12:42,806 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:12:42,807 [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, 5125.95it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4952.65it/s]
INFO:     172.17.0.1:39290 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39290 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39290 - "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.
2026-10-05 04:10:23.940242780 [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]+
[INFO] 2026-10-05 04:10:24,385 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:24,387 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:10:24,554 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:24,555 [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] 2026-10-05 04:10:24,637 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:24,640 [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%|██████▍   | 494/770 [00:00<00:00, 4938.37it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4936.10it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:40996 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:40998 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:10:27,729 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:27,729 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:10:27,865 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:27,865 [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] 2026-10-05 04:10:27,916 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:27,916 [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%|██████▍   | 495/770 [00:00<00:00, 4942.11it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5715.93it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:40998 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:10:31,813 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:31,814 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:10:31,909 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:31,909 [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] 2026-10-05 04:10:31,964 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:31,965 [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%|██████▍   | 500/770 [00:00<00:00, 4910.82it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5091.53it/s]
INFO:     172.17.0.1:40998 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:10:33,844 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:33,844 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:10:33,942 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:33,942 [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] 2026-10-05 04:10:33,991 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:33,991 [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%|██████▏   | 476/770 [00:00<00:00, 4752.92it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4667.03it/s]
INFO:     172.17.0.1:40998 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:40998 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:40998 - "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.
2026-10-05 04:09:56.310815971 [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]+
[INFO] 2026-10-05 04:09:57,917 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:09:57,920 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:09:58,067 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:09:58,068 [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] 2026-10-05 04:09:58,123 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:09:58,123 [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:  79%|███████▊  | 605/770 [00:00<00:00, 6047.25it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 6151.48it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:54024 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:54028 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:10:00,344 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:00,345 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:10:00,437 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:00,438 [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] 2026-10-05 04:10:00,482 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:00,482 [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%|██████▊   | 526/770 [00:00<00:00, 5258.58it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5272.33it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:54028 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:10:02,411 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:02,412 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:10:02,519 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:02,520 [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] 2026-10-05 04:10:02,587 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:02,587 [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:  70%|███████   | 542/770 [00:00<00:00, 5395.05it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5865.47it/s]
INFO:     172.17.0.1:54028 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:10:06,389 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:06,390 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:10:06,473 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:06,473 [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] 2026-10-05 04:10:06,518 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:10:06,518 [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:  69%|██████▊   | 529/770 [00:00<00:00, 5284.36it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5399.45it/s]
INFO:     172.17.0.1:54028 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:54028 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:54028 - "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.
2026-10-05 04:07:23.167702904 [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]+
[INFO] 2026-10-05 04:07:23,976 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:23,979 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:07:24,131 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:24,131 [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] 2026-10-05 04:07:24,179 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:24,180 [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%|█████▌    | 430/770 [00:00<00:00, 4282.36it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4866.19it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:59638 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:59652 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:07:26,207 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:26,207 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:07:26,324 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:26,324 [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] 2026-10-05 04:07:26,399 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:26,399 [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:  55%|█████▍    | 423/770 [00:00<00:00, 4209.19it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4958.42it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:59652 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:07:30,213 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:30,213 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:07:30,299 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:30,299 [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] 2026-10-05 04:07:30,355 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:30,355 [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:  54%|█████▎    | 412/770 [00:00<00:00, 4117.78it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4219.78it/s]
INFO:     172.17.0.1:59652 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:07:32,221 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:32,222 [RapidOCR] main.py:57: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_mobile.onnx
[INFO] 2026-10-05 04:07:32,335 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:32,335 [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] 2026-10-05 04:07:32,386 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:32,387 [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:  55%|█████▌    | 427/770 [00:00<00:00, 4251.16it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4291.93it/s]
INFO:     172.17.0.1:59652 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:59652 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:59652 - "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.
2026-10-05 04:07:41.010900537 [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]+
[INFO] 2026-10-05 04:07:41,639 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:41,640 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 04:07:41,759 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:41,759 [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] 2026-10-05 04:07:41,824 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:41,825 [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:  62%|██████▏   | 478/770 [00:00<00:00, 4726.25it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 4594.55it/s]
INFO:     Application startup complete.
INFO:     Uvicorn running on http://0.0.0.0:5001 (Press CTRL+C to quit)
INFO:     172.17.0.1:55442 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:55454 - "GET /health HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:07:43,874 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:43,874 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 04:07:43,966 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:43,966 [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] 2026-10-05 04:07:44,012 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:44,012 [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:  69%|██████▉   | 531/770 [00:00<00:00, 5309.21it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5334.28it/s]
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:55454 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:07:45,959 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:45,960 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 04:07:46,123 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:46,123 [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] 2026-10-05 04:07:46,177 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:46,177 [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:  56%|█████▌    | 433/770 [00:00<00:00, 4306.16it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5058.19it/s]
INFO:     172.17.0.1:55454 - "POST /v1/convert/source HTTP/1.1" 200 OK
[INFO] 2026-10-05 04:07:50,072 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:50,072 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 04:07:50,172 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:50,172 [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] 2026-10-05 04:07:50,224 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:07:50,225 [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:  69%|██████▉   | 535/770 [00:00<00:00, 5348.98it/s]
Loading weights: 100%|██████████| 770/770 [00:00<00:00, 5540.52it/s]
INFO:     172.17.0.1:55454 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:55454 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:55454 - "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.
2026-10-05 04:04:27.873690182 [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]+
[INFO] 2026-10-05 04:04:28,347 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:04:28,349 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 04:04:28,428 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:04:28,429 [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] 2026-10-05 04:04:28,471 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:04:28,472 [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:     172.17.0.1:36640 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:36656 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:36656 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36656 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36656 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:36656 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:36656 - "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.
2026-10-05 04:04:17.384850719 [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]+
[INFO] 2026-10-05 04:04:19,258 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:04:19,261 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 04:04:19,413 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:04:19,414 [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] 2026-10-05 04:04:19,477 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:04:19,477 [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:     172.17.0.1:50676 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:50680 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:50680 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:50680 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:50680 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:50680 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:50680 - "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.
2026-10-05 04:01:19.686035195 [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]+
[INFO] 2026-10-05 04:01:20,155 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:01:20,157 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 04:01:20,283 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:01:20,284 [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] 2026-10-05 04:01:20,343 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:01:20,343 [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:     172.17.0.1:48698 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:48712 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:48712 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48712 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48712 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48712 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:48712 - "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.
2026-10-05 04:01:07.590524146 [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]+
[INFO] 2026-10-05 04:01:08,309 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:01:08,312 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 04:01:08,474 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:01:08,474 [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] 2026-10-05 04:01:08,543 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 04:01:08,543 [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:     172.17.0.1:43138 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:43144 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:43144 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43144 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43144 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43144 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:43144 - "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.
2026-10-05 03:57:47.198083814 [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]+
[INFO] 2026-10-05 03:57:48,156 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:57:48,160 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:57:48,329 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:57:48,329 [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] 2026-10-05 03:57:48,381 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:57:48,381 [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:     172.17.0.1:38176 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:38188 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:38188 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:38188 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:38188 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:38188 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:38188 - "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.
2026-10-05 03:58:09.962532371 [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]+
[INFO] 2026-10-05 03:58:10,619 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:58:10,621 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:58:10,728 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:58:10,728 [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] 2026-10-05 03:58:10,793 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:58:10,793 [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:     172.17.0.1:48950 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:48962 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:48962 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48962 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48962 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:48962 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:48962 - "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.
2026-10-05 03:54:43.806518051 [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]+
[INFO] 2026-10-05 03:54:44,313 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:54:44,315 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:54:44,431 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:54:44,431 [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] 2026-10-05 03:54:44,478 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:54:44,478 [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:     172.17.0.1:53566 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:53570 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:53570 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53570 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53570 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:53570 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:53570 - "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.
2026-10-05 03:54:33.666208493 [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]+
[INFO] 2026-10-05 03:54:35,704 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:54:35,706 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:54:35,832 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:54:35,833 [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] 2026-10-05 03:54:35,893 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:54:35,894 [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:     172.17.0.1:32830 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:32844 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:32844 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:32844 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:32844 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:32844 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:32844 - "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] 2026-10-05 03:51:25,653 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:51:25,655 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:51:25,739 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:51:25,740 [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] 2026-10-05 03:51:25,776 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:51:25,776 [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:     172.17.0.1:58596 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:58610 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:58610 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58610 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58610 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58610 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:58610 - "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] 2026-10-05 03:51:09,852 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:51:09,854 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:51:11,419 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:51:11,419 [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] 2026-10-05 03:51:11,515 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:51:11,516 [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:     172.17.0.1:39748 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39764 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39764 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39764 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39764 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39764 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39764 - "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] 2026-10-05 03:48:16,910 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:48:16,912 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:48:17,015 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:48:17,015 [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] 2026-10-05 03:48:17,054 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:48:17,055 [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:     172.17.0.1:58966 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:58968 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:58968 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58968 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58968 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58968 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:58968 - "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] 2026-10-05 03:46:57,749 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:46:57,751 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:46:57,870 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:46:57,871 [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] 2026-10-05 03:46:57,929 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:46:57,929 [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:     172.17.0.1:43964 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:43980 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:43980 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43980 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43980 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:43980 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:43980 - "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] 2026-10-05 03:44:31,269 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:44:31,271 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:44:31,371 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:44:31,371 [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] 2026-10-05 03:44:31,419 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:44:31,419 [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:     172.17.0.1:33754 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:33758 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:33758 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:33758 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:33758 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:33758 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:33758 - "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] 2026-10-05 03:42:59,304 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:42:59,306 [RapidOCR] main.py:53: Using /opt/app-root/src/.cache/docling/models/RapidOcr/onnx/PP-OCRv4/det/ch_PP-OCRv4_det_infer.onnx
[INFO] 2026-10-05 03:42:59,425 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:42:59,425 [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] 2026-10-05 03:42:59,511 [RapidOCR] base.py:22: Using engine_name: onnxruntime
[INFO] 2026-10-05 03:42:59,511 [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:     172.17.0.1:33466 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:33468 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:33468 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:33468 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:33468 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:33468 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:33468 - "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:     172.17.0.1:42124 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:42134 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:42134 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:42134 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:42134 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:42134 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:42134 - "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:     172.17.0.1:39004 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39006 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39006 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39006 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39006 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39006 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39006 - "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:     172.17.0.1:44908 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:44910 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:44910 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:44910 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:44910 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:44910 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:44910 - "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:     172.17.0.1:37620 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:37626 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:37626 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:37626 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:37626 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:37626 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:37626 - "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:     172.17.0.1:55652 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:55658 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:55658 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:55658 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:55658 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:55658 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:55658 - "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:     172.17.0.1:54838 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:54840 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:54840 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:54840 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:54840 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:54840 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:54840 - "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:     172.17.0.1:39390 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39396 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39396 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39396 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39396 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39396 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39396 - "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:     172.17.0.1:52856 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:52860 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:52860 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:52860 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:52860 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:52860 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:52860 - "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:     172.17.0.1:39114 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:39122 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:39122 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39122 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39122 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:39122 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:39122 - "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:     172.17.0.1:57694 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:57708 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:57708 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:57708 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:57708 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:57708 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:57708 - "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:     172.17.0.1:44354 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:44370 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:44370 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:44370 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:44370 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:44370 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:44370 - "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:     172.17.0.1:58836 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:58848 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:58848 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58848 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58848 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:58848 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:58848 - "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:     172.17.0.1:54452 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:54454 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:54454 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:54454 - "POST /v1/convert/source 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:     172.17.0.1:49356 - "GET /health HTTP/1.1" 200 OK
INFO:     172.17.0.1:42940 - "GET /health HTTP/1.1" 200 OK
WARNING:docling_core.types.doc.document:Parameter `strict_text` has been deprecated and will be ignored.
INFO:     172.17.0.1:42940 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:42940 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:42940 - "POST /v1/convert/source HTTP/1.1" 200 OK
INFO:     172.17.0.1:42940 - "GET /v1/clear/converters HTTP/1.1" 200 OK
INFO:     172.17.0.1:42940 - "GET /v1/clear/results?older_then=3600 HTTP/1.1" 200 OK