5 results found

1. Explicit Context Caching

   Support for Explicit Context Caching in the Gemini API. When applications frequently sends long, static system instructions and large document sets (100k+ tokens) in every prompt. Implementing an explicit cache would significantly reduce latency for users and lower the token overhead by allowing us to persist this context across multiple turns without re-sending the entire prefix.

   1 vote

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  Firebase AI Logic  ·  Edit…  ·  Delete…  ·  Admin →
   declined  ·  AdminMiguel Ramos (Product Manager @Google, Google, LLC - Firebase) responded

   Training models is not part of the product direction of Firebase AI Logic. Other platforms like Vertex AI are more suitable for that.

2. Train A Model Through API

   I don't know if this is already exisiting but i think it should be possible to train a custom model through the vertex AI API for firebase.

   5 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  Firebase AI Logic  ·  Delete…  ·  Admin →
   declined  ·  AdminMiguel Ramos (Product Manager @Google, Google, LLC - Firebase) responded

   Training models is not part of the product direction of Firebase AI Logic. Other platforms like Vertex AI are more suitable for that.

3. Automated Function Calling for Kotlin

   Allow the SDK to call automatically the functions that the model requires before to proceed with the prompt

   3 votes

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   0 comments  ·  Firebase AI Logic  ·  Delete…  ·  Admin →
   declined  ·  AdminMiguel Ramos (Product Manager @Google, Google, LLC - Firebase) responded

   Training models is not part of the product direction of Firebase AI Logic. Other platforms like Vertex AI are more suitable for that.

4. Authentication error: 7

   Luego de inicial sesion en firebase.com ya no puedo ver mis proyectos. Me tira el error
   Authentication error: 7; Error Details: User not allowed to access GCP services.
   Estoy investigando pero todavía no di con la solucion, si alguien me puede ayudar. Desde ya gracias

   1 vote

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   declined  ·  1 comment  ·  Authentication  ·  Delete…  ·  Admin →

5. Security vulnerability on using insecure function like execSQL which is found in firebase android sdk

   firebase-android-sdk in SchemaManager.java it uses execSQL. This is being raised as as a security vulnerability from the Security Audit. Please find the path below :
   transport/transport-runtime/src/main/java/com/google/android/datatransport/runtime/scheduling/persistence/SchemaManager.java

   Is this will affect the application as we use firebase in our application? Is it a false positive? or in the further version alternative of execSQL can be used.

   1 vote

   Vote Vote Vote Vote

   We're glad you're here

   Please sign in to leave feedback

   Signed in as (Sign out)

   Close

   Close

   Vote

   We’ll send you updates on this idea

   1 comment  ·  Firebase console  ·  Delete…  ·  Admin →
   declined  ·  AdminFrank van Puffelen (Firebaser, Google, LLC - Firebase) responded

   The execSQL statements identified manipulate the local SQLite database that the Android SDK uses to cache Firestore data. They're not a security risk.

   This is also not a feature request. Going forward please report security concerns directly on the relevant repo, rather than here.