import { BaseCache } from '@langchain/core/caches'
import { ChatVertexAIInput, ChatVertexAI as LcChatVertexAI } from '@langchain/google-vertexai'
import { buildGoogleCredentials } from '../../../src/google-utils'
import {
ICommonObject,
IMultiModalOption,
INode,
INodeData,
INodeOptionsValue,
INodeParams,
IVisionChatModal
} from '../../../src/Interface'
import { getModels, getRegions, MODEL_TYPE } from '../../../src/modelLoader'
import { getBaseClasses } from '../../../src/utils'
const DEFAULT_IMAGE_MAX_TOKEN = 8192
const DEFAULT_IMAGE_MODEL = 'gemini-1.5-flash-latest'
class ChatVertexAI extends LcChatVertexAI implements IVisionChatModal {
configuredModel: string
configuredMaxToken: number
multiModalOption: IMultiModalOption
id: string
constructor(id: string, fields?: ChatVertexAIInput) {
// @ts-ignore
if (fields?.model) {
fields.modelName = fields.model
delete fields.model
}
super(fields ?? {})
this.id = id
this.configuredModel = fields?.modelName || ''
this.configuredMaxToken = fields?.maxOutputTokens ?? 2048
}
revertToOriginalModel(): void {
this.modelName = this.configuredModel
this.maxOutputTokens = this.configuredMaxToken
}
setMultiModalOption(multiModalOption: IMultiModalOption): void {
this.multiModalOption = multiModalOption
}
setVisionModel(): void {
if (!this.modelName.startsWith('claude-3')) {
this.modelName = DEFAULT_IMAGE_MODEL
this.maxOutputTokens = this.configuredMaxToken ? this.configuredMaxToken : DEFAULT_IMAGE_MAX_TOKEN
}
}
}
class GoogleVertexAI_ChatModels implements INode {
label: string
name: string
version: number
type: string
icon: string
category: string
description: string
baseClasses: string[]
credential: INodeParams
inputs: INodeParams[]
constructor() {
this.label = 'ChatGoogleVertexAI'
this.name = 'chatGoogleVertexAI'
this.version = 5.3
this.type = 'ChatGoogleVertexAI'
this.icon = 'GoogleVertex.svg'
this.category = 'Chat Models'
this.description = 'Wrapper around VertexAI large language models that use the Chat endpoint'
this.baseClasses = [this.type, ...getBaseClasses(ChatVertexAI)]
this.credential = {
label: 'Connect Credential',
name: 'credential',
type: 'credential',
credentialNames: ['googleVertexAuth'],
optional: true,
description:
'Google Vertex AI credential. If you are using a GCP service like Cloud Run, or if you have installed default credentials on your local machine, you do not need to set this credential.'
}
this.inputs = [
{
label: 'Cache',
name: 'cache',
type: 'BaseCache',
optional: true
},
{
label: 'Region',
description: 'Region to use for the model.',
name: 'region',
type: 'asyncOptions',
loadMethod: 'listRegions',
optional: true
},
{
label: 'Model Name',
name: 'modelName',
type: 'asyncOptions',
loadMethod: 'listModels'
},
{
label: 'Custom Model Name',
name: 'customModelName',
type: 'string',
placeholder: 'gemini-1.5-pro-exp-0801',
description: 'Custom model name to use. If provided, it will override the model selected',
additionalParams: true,
optional: true
},
{
label: 'Temperature',
name: 'temperature',
type: 'number',
step: 0.1,
default: 0.9,
optional: true
},
{
label: 'Allow Image Uploads',
name: 'allowImageUploads',
type: 'boolean',
description:
'Allow image input. Refer to the docs for more details.',
default: false,
optional: true
},
{
label: 'Streaming',
name: 'streaming',
type: 'boolean',
default: true,
optional: true,
additionalParams: true
},
{
label: 'Max Output Tokens',
name: 'maxOutputTokens',
type: 'number',
step: 1,
optional: true,
additionalParams: true
},
{
label: 'Top Probability',
name: 'topP',
type: 'number',
step: 0.1,
optional: true,
additionalParams: true
},
{
label: 'Top Next Highest Probability Tokens',
name: 'topK',
type: 'number',
description: `Decode using top-k sampling: consider the set of top_k most probable tokens. Must be positive`,
step: 1,
optional: true,
additionalParams: true
},
{
label: 'Thinking Budget',
name: 'thinkingBudget',
type: 'number',
description: 'Number of tokens to use for thinking process (0 to disable)',
step: 1,
placeholder: '1024',
optional: true,
additionalParams: true
}
]
}
//@ts-ignore
loadMethods = {
async listModels(): Promise {
return await getModels(MODEL_TYPE.CHAT, 'chatGoogleVertexAI')
},
async listRegions(): Promise {
return await getRegions(MODEL_TYPE.CHAT, 'chatGoogleVertexAI')
}
}
async init(nodeData: INodeData, _: string, options: ICommonObject): Promise {
const temperature = nodeData.inputs?.temperature as string
const modelName = nodeData.inputs?.modelName as string
const customModelName = nodeData.inputs?.customModelName as string
const maxOutputTokens = nodeData.inputs?.maxOutputTokens as string
const topP = nodeData.inputs?.topP as string
const cache = nodeData.inputs?.cache as BaseCache
const topK = nodeData.inputs?.topK as string
const streaming = nodeData.inputs?.streaming as boolean
const thinkingBudget = nodeData.inputs?.thinkingBudget as string
const region = nodeData.inputs?.region as string
const allowImageUploads = nodeData.inputs?.allowImageUploads as boolean
const multiModalOption: IMultiModalOption = {
image: {
allowImageUploads: allowImageUploads ?? false
}
}
const obj: ChatVertexAIInput = {
temperature: parseFloat(temperature),
modelName: customModelName || modelName,
streaming: streaming ?? true
}
const authOptions = await buildGoogleCredentials(nodeData, options)
if (authOptions && Object.keys(authOptions).length !== 0) obj.authOptions = authOptions
if (maxOutputTokens) obj.maxOutputTokens = parseInt(maxOutputTokens, 10)
if (topP) obj.topP = parseFloat(topP)
if (cache) obj.cache = cache
if (topK) obj.topK = parseFloat(topK)
if (thinkingBudget) obj.thinkingBudget = parseInt(thinkingBudget, 10)
if (region) obj.location = region
const model = new ChatVertexAI(nodeData.id, obj)
model.setMultiModalOption(multiModalOption)
return model
}
}
module.exports = { nodeClass: GoogleVertexAI_ChatModels }