You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Not convinced that we need to do this, but we can spare a few bytes in the NLU results files by only including the timestamp when requested via CLI argument.
Copy file name to clipboardExpand all lines: docs/Test.md
+5Lines changed: 5 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -140,6 +140,11 @@ See [Caching speech-to-text transcriptions](#caching-speech-to-text-transcriptio
140
140
141
141
This is currently only used for LUIS, see the section on LUIS prebuilt entities in [Configuring prebuilt entities](LuisModelConfiguration.md#configuring-prebuilt-entities).
142
142
143
+
### `--timestamp`
144
+
(Optional) Signals whether to add a timestamp to each NLU test result.
145
+
146
+
See the documentation on the [`timestamp` property](UtteranceExtensions.md#returning-timestamps-for-each-query) for more details.
147
+
143
148
### `-i, --include`
144
149
(Optional) Path to custom NLU provider DLL. See documentation about [Specifying the include path](https://github.com/microsoft/NLU.DevOps/blob/master/docs/CliExtensions.md#specifying-the-include-path) for more details.
Copy file name to clipboardExpand all lines: docs/UtteranceExtensions.md
+2-2Lines changed: 2 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -22,7 +22,7 @@ When an NLU provider in NLU.DevOps returns a prediction result, the value will b
22
22
```
23
23
In this case, the intent confidence score was `0.99` and the text transcription confidence score was `0.95`. This is useful context when debugging false predictions, as a low confidence score may indicate that the model could be improved with more training examples. The recognized `genre` entity also includes a confidence score of `0.80`, although it should be noted that only the LUIS provider currently returns confidence score for entity types trained from examples.
24
24
25
-
## Labeled utterance timestamps
25
+
## Returning timestamps for each query
26
26
27
27
When analyzing results for a set of NLU predictions, it is often important context to understand when the test was run. For example, for Dialogflow `date` and `time` entities, the service only returns a date time string, and no indication of what token(s) triggered that entity to be recognized. For example, the result from a query like `"Call a taxi in 15 minutes"` may look like the following:
28
28
```json
@@ -38,7 +38,7 @@ When analyzing results for a set of NLU predictions, it is often important conte
38
38
"timestamp": "2020-01-01T00:00:00-04:00"
39
39
}
40
40
```
41
-
Without the context provided by the `timestamp` property, we wouldn't be able to make any assertion about the correctness of the `entityValue` property for time. Currently, LUIS, Lex, and Dialogflow return a timestamp for each prediction result.
41
+
Without the context provided by the `timestamp` property, we wouldn't be able to make any assertion about the correctness of the `entityValue` property for time. Currently, you must specify the [`--timestamp`](Test.md#--timestamp) option to ensure a timestamp is assigned to each NLU prediction result.
0 commit comments