Semantic search token usage tracking #63591
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
lbrdnk
force-pushed
the
sem-search-token-tracking
branch
from
3b8e11b to
f39c72d
Compare
lbrdnk
force-pushed
the
sem-search-token-tracking
branch
from
f39c72d to
ba98ce2
Compare
lbrdnk
added
the
backport
lbrdnk
changed the title
e2e tests failed on
|
| File | Test Name |
|---|---|
embedding-reproductions.cy.spec.js |
(flaky) issue 40660 > static dashboard content shouldn't overflow its container (#40660) |
piranha
reviewed
Sep 23, 2025
| constraints: | ||
| nullable: false | ||
| - column: | ||
| name: total_tokens |
There was a problem hiding this comment.
do we not want to store in/out separately? :)
There was a problem hiding this comment.
@piranha probably no need as it's an embedding model, so tracking total tokens is enough
piranha
approved these changes
Sep 23, 2025
Comment on lines
+282
to
+286
| (is (= 1 (t2/count :model/SemanticSearchTokenTracking))) | ||
| (let [{:keys [request_type total_tokens]} | ||
| (t2/select-one :model/SemanticSearchTokenTracking)] | ||
| (is (= :index request_type)) | ||
| (is (= 13 total_tokens)))) |
There was a problem hiding this comment.
This is by no means a request to change, but I personally find it easier to read =? stuff, like this:
(is (=? [{:request_type :index :total_tokens 13}]
(t2/select :model/SemanticSearchTokenTracking)))
github-automation-metabase
pushed a commit
that referenced
this pull request
Sep 23, 2025
* Pass type of embedding request * Initial migration * Add remaining opts args * Add SemanticSearchTokenTracking module * Connect token tracking to ai-service impl * Add test for token tracking writes. * Remove prompt_tokens including migration * Add usage trimmer job * Trimmer test * Activate trimmer job * test * Record tokens for openai * Update test * Use total-tokens * Comment * linter * Add index * Exclude from copy
metamben
pushed a commit
that referenced
this pull request
Sep 23, 2025
* Pass type of embedding request * Initial migration * Add remaining opts args * Add SemanticSearchTokenTracking module * Connect token tracking to ai-service impl * Add test for token tracking writes. * Remove prompt_tokens including migration * Add usage trimmer job * Trimmer test * Activate trimmer job * test * Record tokens for openai * Update test * Use total-tokens * Comment * linter * Add index * Exclude from copy
github-automation-metabase
added a commit
that referenced
this pull request
Sep 23, 2025
* Pass type of embedding request * Initial migration * Add remaining opts args * Add SemanticSearchTokenTracking module * Connect token tracking to ai-service impl * Add test for token tracking writes. * Remove prompt_tokens including migration * Add usage trimmer job * Trimmer test * Activate trimmer job * test * Record tokens for openai * Update test * Use total-tokens * Comment * linter * Add index * Exclude from copy Co-authored-by: lbrdnk <lbrdnk@users.noreply.github.com>
galdre
pushed a commit
that referenced
this pull request
Sep 24, 2025
* Pass type of embedding request * Initial migration * Add remaining opts args * Add SemanticSearchTokenTracking module * Connect token tracking to ai-service impl * Add test for token tracking writes. * Remove prompt_tokens including migration * Add usage trimmer job * Trimmer test * Activate trimmer job * test * Record tokens for openai * Update test * Use total-tokens * Comment * linter * Add index * Exclude from copy
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Closes https://linear.app/metabase/issue/BOT-410/implement-token-usage-tracking-for-semantic-search
This PR adds tracking of tokens consumed by embeddings api calls initiated from semantic search. Number of tokens of every request is stored in appdb. The tables is trimmed every day, storing rolling 2 months of data.
This PR
:model/SemanticSearchTokenTracking,get-embedding...implementations to take opts,upsert-index!,query-index) with appropriate request:type, that is stored with token count in the new table,