Closed
Description
Is your feature request related to a problem? Please describe.
I've gotten a few reports of folks running into this:
If you write code like
import { geminiPro } from "@genkit-ai/vertexai";
...
const result = await generate({
model: geminiPro,
prompt: `Tell me a joke about {input}`,
output: z.string()
});
without ever calling configureGenkit() to configure Vertex you get and error like
Error: Model {"name":"vertexai/gemini-1.0-pro","info":{"label":"Vertex AI - Gemini Pro","versions":["gemini-1.0-pro","gemini-1.0-pro-001"],"supports":{"multiturn":true,"media":false,"tools":true,"systemRole":true}},"configSchema":{"_def":{"unknownKeys":"strip","catchall":{"_def":{"typeName":"ZodNever"}},"typeName":"ZodObject"},"_cached":null}} not found
at /workspace/node_modules/@genkit-ai/ai/lib/generate.js:480:13
at Generator.next (<anonymous>)
at fulfilled (/workspace/node_modules/@genkit-ai/ai/lib/generate.js:53:24)
at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
Describe the solution you'd like
It would be better if the error said something like "Are you missing a call configureGenkit(...)"
Describe alternatives you've considered
Or if there was a way to throw a error("Called without configuration") that could also be helpful.
Additional context
Add any other context or screenshots about the feature request here.
Metadata
Metadata
Assignees
Labels
Type
Projects
Status