Replies: 12 comments 16 replies
-
|
I'm not sure you are calling the function properly (even though I haven't seen your code). Here is the proper way to do this. Just enter in your semantic model name and workspace name in the corresponding parameters. You can also specify a measure or list of measures using the 'measure_name' parameter within the generate_measure_descriptions function. from sempy_labs.tom import connect_semantic_model
with connect_semantic_model(dataset='', workspace='', readonly=False) as tom:
tom.generate_measure_descriptions() |
Beta Was this translation helpful? Give feedback.
-
|
Yeah thank you I closed the discussion because I found an error in my script. |
Beta Was this translation helpful? Give feedback.
-
|
Great! Happy holidays! |
Beta Was this translation helpful? Give feedback.
-
|
Just a last question: I run the script for all measures in my Dataset and, after few minutes, got this error: FabricHTTPException: 401 Unauthorized for url: https://api.fabric.microsoft.com//explore/v202304/nl2nl/completions I tried the script for a specific measure (added in tom.generate_measure_descriptions(measure_name='mymeasure')) and it worked well; I re-run the same script but got the above error. |
Beta Was this translation helpful? Give feedback.
-
|
You need to run this on a paid F64 sku or higher. |
Beta Was this translation helpful? Give feedback.
-
|
The workspace is assigned to an F512 capacity |
Beta Was this translation helpful? Give feedback.
-
|
Hmm, interesting. I'll have to check and get back to you - likely after the holidays. |
Beta Was this translation helpful? Give feedback.
-
|
This should be resolved in 0.9.0. |
Beta Was this translation helpful? Give feedback.
-
|
hi m-Kovalsky FabricHTTPException: 401 Unauthorized for url: https://msitapi.fabric.microsoft.com//explore/v202304/nl2nl/completions We are executing this in F256 capacity |
Beta Was this translation helpful? Give feedback.
-
|
@m-kovalsky - any update on this |
Beta Was this translation helpful? Give feedback.
-
|
Which region are you using? |
Beta Was this translation helpful? Give feedback.
-
|
West us
…On Wed, Apr 9, 2025, 11:46 PM m-kovalsky ***@***.***> wrote:
Which region are you using?
—
Reply to this email directly, view it on GitHub
<#383 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAEYBVVI5TX2VKIEETT5FN32YYHS5AVCNFSM6AAAAABUETRAXCVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTENZYG42DINY>
.
You are receiving this because you commented.Message ID:
<microsoft/semantic-link-labs/repo-discussions/383/comments/12787447@
github.com>
|
Beta Was this translation helpful? Give feedback.
-
|
Do you get this error immediately or after the function runs for some time? |
Beta Was this translation helpful? Give feedback.
-
|
@m-kovalsky - We are getting this error after 2mins FabricHTTPException: 500 Internal Server Error for url: https://msitapi.fabric.microsoft.com//explore/v202304/nl2nl/completions |
Beta Was this translation helpful? Give feedback.
-
|
this the code we are using it from sempy_labs.tom import connect_semantic_model dataset = "" # Enter the name or ID of your semantic model with connect_semantic_model(dataset=dataset, workspace=workspace, readonly=readonly) as tom: |
Beta Was this translation helpful? Give feedback.
-
|
Ah so that’s a different error. I think I know the issue for that. Do you have measures which have DAX that contains double quotes? |
Beta Was this translation helpful? Give feedback.
-
|
|
Beta Was this translation helpful? Give feedback.
-
|
@m-kovalsky - any solution for this? |
Beta Was this translation helpful? Give feedback.
-
|
So I tested for measures with double quotes and they work fine. Are you able to identify the measure which causes the issue? You can specify a measure or list of measures within the function’s parameter. Sharing the DAX of the problem measure would help a lot. |
Beta Was this translation helpful? Give feedback.
-
|
it has more the 100+ measures. Is there is a way to find which DAX is not supported while using your code Error Log: |
Beta Was this translation helpful? Give feedback.
-
|
Could you try running this? It gets descriptions for 5 measures at a time. Before it runs the API to get the descriptions, it prints out the 5 measure names. Find the last printout of the measure names and within those you will find the culprit. Would then be great to see the DAX for those 5 measures (you can then run those 5 measures individually to find the specific culprit). from sempy_labs.tom import connect_semantic_model
import pandas as pd
from typing import Optional, List
import sempy_labs._icons as icons
dataset = ''
workspace = ''
with connect_semantic_model(dataset=dataset, workspace=workspace, readonly=True) as tom:
def generate_measure_descriptions(
measure_name: Optional[str | List[str]] = None,
max_batch_size: Optional[int] = 5,
) -> pd.DataFrame:
"""
Auto-generates descriptions for measures using an LLM. This function requires a paid F-sku (Fabric) of F64 or higher.
Setting the 'readonly' parameter in connect_semantic_model to True will allow you to see the auto-generated descriptions in a dataframe. Setting the 'readonly' parameter
to False will update the descriptions for the measures within the 'measure_name' parameter.
Parameters
----------
measure_name : str | List[str], default=None
The measure name (or a list of measure names).
Defaults to None which generates descriptions for all measures in the semantic model.
max_batch_size : int, default=5
Sets the max batch size for each API call.
Returns
-------
pandas.DataFrame
A pandas dataframe showing the updated measure(s) and their new description(s).
"""
df = pd.DataFrame(
columns=["Table Name", "Measure Name", "Expression", "Description"]
)
data = []
# import concurrent.futures
if measure_name is None:
measure_name = [m.Name for m in tom.all_measures()]
if isinstance(measure_name, str):
measure_name = [measure_name]
if len(measure_name) > max_batch_size:
measure_lists = [
measure_name[i : i + max_batch_size]
for i in range(0, len(measure_name), max_batch_size)
]
else:
measure_lists = [measure_name]
# Each API call can have a max of 5 measures
for measure_list in measure_lists:
payload = {
"scenarioDefinition": {
"generateModelItemDescriptions": {
"modelItems": [],
},
},
"workspaceId": tom._workspace_id,
"artifactInfo": {"artifactType": "SemanticModel"},
}
for m_name in measure_list:
expr, t_name = next(
(ms.Expression, ms.Parent.Name)
for ms in tom.all_measures()
if ms.Name == m_name
)
if t_name is None:
raise ValueError(
f"{icons.red_dot} The '{m_name}' measure does not exist in the '{tom._dataset_name}' semantic model within the '{tom._workspace_name}' workspace."
)
new_item = {
"urn": m_name,
"type": 1,
"name": m_name,
"expression": expr,
}
payload["scenarioDefinition"]["generateModelItemDescriptions"][
"modelItems"
].append(new_item)
print(measure_list)
response = _base_api(
request="/explore/v202304/nl2nl/completions",
method="post",
payload=payload,
)
for item in response.json().get("modelItems", []):
ms_name = item["urn"]
if ms_name.startswith("urn: "):
ms_name = ms_name[5:]
desc = item.get("description")
(table_name, expr) = next(
(m.Parent.Name, m.Expression)
for m in tom.all_measures()
if m.Name == ms_name
)
tom.model.Tables[table_name].Measures[ms_name].Description = desc
# Collect new descriptions in a dataframe
new_data = {
"Table Name": table_name,
"Measure Name": ms_name,
"Expression": expr,
"Description": desc,
}
data.append(new_data)
if data:
df = pd.concat([df, pd.DataFrame(data)], ignore_index=True)
return df
x = generate_measure_descriptions() |
Beta Was this translation helpful? Give feedback.
-
|
@m-kovalsky with this logic i checked where ever we have "" this code is not working. In real world scenario this will not workout. If you can provide some alternate solutions will be good to use. |
Beta Was this translation helpful? Give feedback.
-
|
That’s odd as I tried this and it works for me. Would you be able to send one of the DAX statements which give an error? |
Beta Was this translation helpful? Give feedback.
-
|
VAR tVar = if possible we can connect to rectify this. That would be more helpful to implement across the portfolio |
Beta Was this translation helpful? Give feedback.
-
|
I cannot reproduce this error. I made a measure with essentially the same code and it works just fine. In which environment are you running this (prod/msit etc.)? And in which region?
|
Beta Was this translation helpful? Give feedback.
-
|
@m-kovalsky using msit and west us |
Beta Was this translation helpful? Give feedback.
-
|
In msit
…On Mon, May 12, 2025, 1:03 AM m-kovalsky ***@***.***> wrote:
I cannot reproduce this error. I made a measure with essentially the same
code and it works just fine. In which environment are you running this
(prod/msit etc.)? And in which region?
image.png (view on web)
<https://github.com/user-attachments/assets/29f84800-f38a-47f0-b52a-1ec3476e6cf8>
—
Reply to this email directly, view it on GitHub
<#383 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AAEYBVTUSGVWCLEDH7YDQOL26BIWBAVCNFSM6AAAAABUETRAXCVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMJRGQ3DKNA>
.
You are receiving this because you commented.Message ID:
<microsoft/semantic-link-labs/repo-discussions/383/comments/13114654@
github.com>
|
Beta Was this translation helpful? Give feedback.
-
|
In msit and west us
…On Mon, May 12, 2025, 1:59 PM sabarinathan R ***@***.***> wrote:
In msit
On Mon, May 12, 2025, 1:03 AM m-kovalsky ***@***.***> wrote:
> I cannot reproduce this error. I made a measure with essentially the same
> code and it works just fine. In which environment are you running this
> (prod/msit etc.)? And in which region?
>
> image.png (view on web)
> <https://github.com/user-attachments/assets/29f84800-f38a-47f0-b52a-1ec3476e6cf8>
>
> —
> Reply to this email directly, view it on GitHub
> <#383 (reply in thread)>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/AAEYBVTUSGVWCLEDH7YDQOL26BIWBAVCNFSM6AAAAABUETRAXCVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMJRGQ3DKNA>
> .
> You are receiving this because you commented.Message ID:
> <microsoft/semantic-link-labs/repo-discussions/383/comments/13114654@
> github.com>
>
|
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi,
I've firstly run the module "connect_semantic_model" in order to connect to a semantic model and being able to use TOM package.
After that, I've run the "labs.tom.TOMWrapper.generate_measure_descriptions" command but obtained this error:
TypeError: TOMWrapper.generate_measure_descriptions() missing 1 required positional argument: 'self'
What is the missing argument? Based on documentation, no argument is mandatory (measure_name and max_batch_size are optional)
Beta Was this translation helpful? Give feedback.
All reactions