What this category's data looks like
Teacher qualification examination and certification materials mainly consist of written and interview registration forms, Mandarin proficiency certificates, and medical examination forms, sourced from kindergarten archives and candidate submissions, and updated by batch. The registration form contains name, ID number, subjects applied for, date, and signature; the Mandarin proficiency certificate contains grade, number, and issue date; the medical examination form contains the medical examination conclusion, date, and physician signature. After materials enter parsing and training tasks, each becomes an independent data row; the registration form, certificate, and medical examination form are each bound to a unique data identifier, split according to chunkSize and chunkSplitter, and the training type trainingType determines the vector or QA path. The field characteristic is that the name, quantity, and date of the same record need to be aligned across materials, and signatures and receipts are indispensable.
Configuration and implementation checks
Parsing and training tasks (server-side tasks) only output data rows and index status (parsed, indexing, indexed); the basis entries that can be traced back to the original text must be handled next by retrieval and subsequent processing. After parsing is completed, parsing and chunking are immediately persisted as visible data rows and bound to a unique data identifier; the index transitions through three states, and data pending indexing does not participate in recall.
| Object | Check | Verification action |
|---|---|---|
| Written and interview registration form | Fields contain name, ID number, subjects applied for, date, and signature, and must be aligned item by item with the Mandarin proficiency certificate and medical examination form | Open the data row to check whether the status is parsed / indexing / indexed, and trace back to the original text to verify the signature and date |
| Mandarin proficiency certificate | Grade, number, and issue date must be consistent with the name and ID number on the registration form, and the version must be current | Verify the file record and the data row separately, confirm that the identifiers can be distinguished, and check whether the version is synchronized |
| Medical examination form | The medical examination conclusion, date, and physician signature must be complete; missing signature or receipt means inconsistency | View the data rows covered by the rebuild task and confirm that only data that has completed indexing is covered |
| Data row | Is in one of the three states and is consistent with the interface display; failed training data can be edited or deleted | Check chunkSize, chunkSplitter, trainingType, and index status |
Three common errors
- Symptom: The name and number of the same record do not match between the registration form and the Mandarin proficiency certificate. Cause: The versions of the two materials were not synchronized or were not compared item by item during entry.
- Symptom: The medical examination form is missing the physician signature or receipt, and the date is inconsistent with the registration form. Cause: Signatures were not fully collected during material circulation, and field verification was not performed during archiving.
- Symptom: Newly uploaded materials cannot be found, and empty data rows remain after training fails. Cause: Successful parsing does not equal retrieval hit; the index status must also be checked, and data pending indexing does not participate in recall.
How to verify the setup
This feature acceptance: Upload one material and observe the data row status transitions and the coverage scope of the rebuild task. The status should be parsed / indexing / indexed; rebuild only covers data that has completed indexing, and the initial training covers data pending indexing and indexing.
Business test: Ask a basis question across materials, see whether the answer provides a traceable source, and verify that the tokens of the quoted text and the number of array entries are recorded separately.
- Open the data row to check whether the status is consistent with the interface display, and confirm that the identities of the file record and the data row can be distinguished from each other.
- View the data rows covered by the rebuild task, and confirm that only data that has completed indexing is covered, and data pending indexing does not participate in the rebuild chain.