Clarity on Model Naming Conventions

[ydax]

[READ] Step 1: Are you in the right place?

Issues filed here should be about bugs for a specific extension in this repository. If you have a general question, need help debugging, or fall into some other category use one of these other channels:

• For general technical questions, post a question on StackOverflow with the firebase tag.
• For general Firebase discussion, use the firebase-talk google group.
• To file a bug against the Firebase Extensions platform, or for an issue affecting multiple extensions, please reach out to Firebase support directly.

[REQUIRED] Step 2: Describe your configuration

• Extension name: _ Multimodal Tasks with the Gemini API
• Extension version: _ 1
• Configuration values (redact info where appropriate):
  • _ Gemini model: gemini-pro

[REQUIRED] Step 3: Describe the problem

The extension doesn't seem to use the Vertex model naming conventions. For example, gemini-pro, the default model name, is nowhere to be found in the linked models docs. gemini-1.0-pro is there and so is gemini-1.5-pro, but no gemini-pro. So which model are we getting? How can we ensure use of, for example, gemini-1.5-pro?

These naming conventions need to line up.

Expected result

I'd recommend that the parameter value for "Gemini model" for this extension line up with the API model names like gemini-1.5-pro that are listed in the docs so that devs can clearly select their preferred model.

[pr-Mais]

Hi @ydax, thanks for your valuable report. The extension allows free-text form for the model name, the one you mentioned gemini-pro is only a default value that you can change, there's no other specified models inside the extension. Linked a PR to update the default value to gemini-1.0-pro which supports the latest stable version of Gemini 1.0 on both Google AI and Vertex AI.

[ydax]

awesome, thanks for clarity. So does that mean that we can use the other model names as seen in the Vertex docs? I tried these model names before and the extension threw an error.

[pr-Mais]

@ydax absolutely, you can put any of the models available in Vertex AI. The extension throwing may indicate an issue, could you let me know the error you were seeing?

[mrroccodev]

Hi @pr-Mais . I have tried to use these model names and got this error: gemini-1.5-flash gemini-flash

What are the list of available models supported?

GoogleGenerativeAIError: [404 Not Found] models/gemini-flash is not found for API version v1, or is not supported for GenerateContent. Call ListModels to see the list of available models and their supported methods. at makeRequest (/workspace/node_modules/@google/generative-ai/dist/index.js:216:19)

[pr-Mais]

@mrroccodev thanks for reporting your issue, we would need to update the API version used by the extension to support newer models & document the supported list, working on it.