Usage
Endpointsc
Once you have installed the Ai package, new endpoint will be available in your project.
Documentation will also be added to your project for full details on each endpoint.
Chat
The chat endpoint allows you to interact with the AI in a chat like manner.
//app.service.js
import { AiService } from '@juicyllama/ai'
@Injectable()
export class AppService {
constructor(private readonly aiService: AiService) {}
async chat(question: string) {
const openaiOptions: CreateChatCompletionRequest = {
model: 'text-davinci-003',
}
const result = await this.aiService.chat({
question: question,
openaiOptions: openaiOptions,
})
if (result.success === AiSuccessType.SUCCESS) {
return result.response
} else {
throw new Error(`[${result.success}] ${result.error_message}`)
}
}
}
Image
The image endpoint allows you to interact with the AI to generate an image.
//app.service.js
import { AiService } from '@juicyllama/ai'
@Injectable()
export class AppService {
constructor(private readonly aiService: AiService) {}
async image(image_description: string) {
const openaiOptions: CreateImageRequest = {}
const result = await this.aiService.image({
image_description: image_description,
openaiOptions: openaiOptions,
})
if (result.success === AiSuccessType.SUCCESS) {
return result.response
} else {
throw new Error(`[${result.success}] ${result.error_message}`)
}
}
}