Index of every page on this site: llms.txt
All skills
microsoft avatar

/azure-ai

@0d60658
by microsoftmicrosoft/skills3.2k stars
373
Run it yourself

Build and troubleshoot Azure AI Search, Speech, Document Intelligence, Content Understanding, Translator, Content Safety, and OpenAI apps. WHEN: AI Search, vector search, hybrid search, semantic search, indexer, speech-to-text, text-to-speech, fast or batch transcription, Speech SDK logs, session ID, result ID, OCR, document translation, glossary.

Use this Skill: https://skilld.dev/gh/microsoft/skills/azure-ai

Nothing lands on disk. Nothing to clean up.

Fork this Skill

Edit a local copy. It keeps the author and licence.

referencesdocument-intelligencecustom-models.md

≈1.5k tokens on demand. Your agent reads this file only when SKILL.md points to it.

Document Intelligence custom models: labels, confidence, training and composition

For custom extraction (template or neural), custom classification and composed models on v4.0 (API 2024-11-30): fields returned empty or with unexpected confidence, accuracy that dropped after retraining, dataset design, classifier or composed-model routing, and copying models. Input files and label-file syntax are in preflight and diagnostics; storage and Studio permissions in Studio and storage.

Read confidence correctly

  • Field confidence is an estimated probability (0–1) and reflects confidence in the position of the extracted value. Neural models report no model accuracy score.
  • Not all fields return a confidence score: a missing confidence is unknown, not 0.
  • Learn (accuracy-confidence, "Table, row, and cell confidence" FAQ): "You should expect an empty field with high confidence for missing values that are mostly empty in the training set too." A high-confidence empty field is usually a training-data signal, not proof that the value is absent.

Diagnose an empty or wrong field

  1. Audit label coverage offline. Download the training container to a local folder (az storage blob download-batch --auth-mode login -s <container> -d <folder>), then run the training-set auditor (Bash / PowerShell; no Node: manual counts):

    node "<skill-directory>/scripts/audit-di-training-set.mjs" --dir "<local-training-folder>"
    & node "<skill-directory>\scripts\audit-di-training-set.mjs" --dir "<local-training-folder>"

    It reports, per field, how many documents carry a labeled value overall and per file type, fields never labeled, label names missing from fields.json, missing .ocr.json files and OCR apiVersion drift. It prints counts and field names only (--anonymize-fields hides names). A field labeled in some documents but not others needs review: confirm the value is truly absent in the unlabeled ones. The launcher exit codes are 0 report/help, 1 folder/read/JSON error, 2 usage error and 3 Node.js 22+ unavailable.

  2. Same input, two models. Analyze the same file with the previous and the new model. If the layout text is the same but fields differ, the cause is the model or its training data. Compare field presence and confidence, never values.

  3. Compare model metadata: GET {endpoint}/documentintelligence/documentModels/{modelId}?api-version=2024-11-30 returns buildMode, apiVersion, createdDateTime, expirationDateTime, trainingHours and docTypes.*.fieldSchema (and fieldConfidence where provided).

Dataset rules from Learn

  • At least five labeled documents; for neural models at least five labeled samples for each variation. About 15 or fewer variations: one model with samples of each; more: separate models and compose them.
  • Include documents where optional fields are present. Label consistently: pick the same context for a value across documents; label the value, not its caption.
  • Neural models don't recognize values split across page boundaries. Signature fields need at least five labeled samples. A token can be labeled for at most two overlapping fields, using region labeling.
  • Retraining with a newer API version needs .ocr.json files produced by that version (analyzeResult.apiVersion). Models built with a GA API expire after two years.
  • Neural training time is capped by maxTrainingHours and billed with a 30-minute minimum per job. Learn conflict on free training: "10 hours of free model training" vs "over 20 requests in a calendar month are billed".

Classify first, then extract

  • Custom classifier: 2–1,000 classes, 5–100 samples per class; splitMode defaults to none in 2024-11-30; handle unknown documents with a score threshold or an other class; incremental training uses baseClassifierId with the same API version.
  • Composed model (v4): needs an explicit classifier, up to 500 component models (F0: 5), optional confidenceThreshold; unmapped document types are ignored. Classification is billed on all pages, extraction only on routed pages.
  • Two calls (classify, then the chosen extraction model) give full control; a composed model routes in one call.

Copy and migrate

authorizeCopy on the target resource, then copyTo on the source with the unchanged payload (valid 24 hours), then poll. Copying between encrypted and unencrypted resources needs x-ms-forms-copy-degrade: true. Copy moves the model, not the training data.

Sources (retrieved 2026-11-18): Accuracy and confidence · Labeled datasets · Label tips · Custom neural · Model lifecycle · Custom classifier · Composed models · Disaster recovery (copy)

Source: SKILL.md on GitHub

skilld matched fixed text patterns in SKILL.md and file names. Patterns miss obfuscated code.

skilld run checks every file with the same patterns. It asks for approval before it loads a Skill with a behavior marked Needs approval.

1 warning11d4 checks · Risk SAFE
  • Gen Agent Trust Hub11d

    This skill provides tools and guidance for developing and troubleshooting Azure AI services. It includes a local diagnostic script for analyzing Speech SDK logs with robust safety controls, along with comprehensive references for official Azure SDKs and secure authentication practices.

  • Socket11d

    No alerts

  • Snyk11d

    Risk: LOW · No issues

  • Runlayer9mo

    15/15 files flagged

Signed by skilld at 0d60658. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated 2 days ago
metadata
{
  "author": "Microsoft",
  "version": "1.2.6"
}

README badge

README badge for microsoft/skills/azure-ai