Andreas/LibreChat
1
0
Fork 0
mirror of https://github.com/danny-avila/LibreChat.git synced 2026-02-22 10:24:09 +01:00
Code Issues Projects Releases Packages Wiki Activity Actions 3

📦 fix: npm warnings; chore: bump deprecated packages (#4707)

Browse source 
* chore: bump langchain deps to address vulnerability warnings

* chore: bump community package and install textsplitters package

* fix: update expected result in tokenSplit tests for accuracy

* chore: remove CodeSherpa tools

* chore: remove E2B tools and loadToolSuite

* chore: remove CodeBrew tool and update related references

* chore: remove HumanTool and ChatTool, update tool references

* chore: remove Zapier tool from manifest.json and update SerpAPI

* chore: remove basic tools

* chore: update import path for RecursiveCharacterTextSplitter

* chore: update import path for DynamicStructuredTool

* chore: remove extractionChain.js and update tool filtering logic

* chore: npm audit fix

* chore: bump google packages

* chore: update DALL-E tool to DALL-E-3 and adjust authentication logic

* ci: update message classes

* chore: elliptic npm audit fix

* chore: update CallbackManager import and remove deprecated tool handling logic

* chore: imports order

* chore: remove unused code

---------

Co-authored-by: Max Sanna <max@maxsanna.com>
This commit is contained in:
Danny Avila 2024-11-12 18:51:32 -05:00 • committed by GitHub
parent d012da0065
commit 95201908e9
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
40 changed files with 1446 additions and 3109 deletions
Download patch file Download diff file Expand all files Collapse all files

122
api/app/clients/agents/Functions/FunctionsAgent.js
View file

@ -1,122 +0,0 @@
const { Agent } = require('langchain/agents');
const { LLMChain } = require('langchain/chains');
const { FunctionChatMessage, AIChatMessage } = require('langchain/schema');
const {
ChatPromptTemplate,
MessagesPlaceholder,
SystemMessagePromptTemplate,
HumanMessagePromptTemplate,
} = require('langchain/prompts');
const { logger } = require('~/config');
const PREFIX = 'You are a helpful AI assistant.';
function parseOutput(message) {
if (message.additional_kwargs.function_call) {
const function_call = message.additional_kwargs.function_call;
return {
tool: function_call.name,
toolInput: function_call.arguments ? JSON.parse(function_call.arguments) : {},
log: message.text,
};
} else {
return { returnValues: { output: message.text }, log: message.text };
}
}
class FunctionsAgent extends Agent {
constructor(input) {
super({ ...input, outputParser: undefined });
this.tools = input.tools;
}
lc_namespace = ['langchain', 'agents', 'openai'];
_agentType() {
return 'openai-functions';
}
observationPrefix() {
return 'Observation: ';
}
llmPrefix() {
return 'Thought:';
}
_stop() {
return ['Observation:'];
}
static createPrompt(_tools, fields) {
const { prefix = PREFIX, currentDateString } = fields || {};
return ChatPromptTemplate.fromMessages([
SystemMessagePromptTemplate.fromTemplate(`Date: ${currentDateString}\n${prefix}`),
new MessagesPlaceholder('chat_history'),
HumanMessagePromptTemplate.fromTemplate('Query: {input}'),
new MessagesPlaceholder('agent_scratchpad'),
]);
}
static fromLLMAndTools(llm, tools, args) {
FunctionsAgent.validateTools(tools);
const prompt = FunctionsAgent.createPrompt(tools, args);
const chain = new LLMChain({
prompt,
llm,
callbacks: args?.callbacks,
});
return new FunctionsAgent({
llmChain: chain,
allowedTools: tools.map((t) => t.name),
tools,
});
}
async constructScratchPad(steps) {
return steps.flatMap(({ action, observation }) => [
new AIChatMessage('', {
function_call: {
name: action.tool,
arguments: JSON.stringify(action.toolInput),
},
}),
new FunctionChatMessage(observation, action.tool),
]);
}
async plan(steps, inputs, callbackManager) {
// Add scratchpad and stop to inputs
const thoughts = await this.constructScratchPad(steps);
const newInputs = Object.assign({}, inputs, { agent_scratchpad: thoughts });
if (this._stop().length !== 0) {
newInputs.stop = this._stop();
}
// Split inputs between prompt and llm
const llm = this.llmChain.llm;
const valuesForPrompt = Object.assign({}, newInputs);
const valuesForLLM = {
tools: this.tools,
};
for (let i = 0; i < this.llmChain.llm.callKeys.length; i++) {
const key = this.llmChain.llm.callKeys[i];
if (key in inputs) {
valuesForLLM[key] = inputs[key];
delete valuesForPrompt[key];
}
}
const promptValue = await this.llmChain.prompt.formatPromptValue(valuesForPrompt);
const message = await llm.predictMessages(
promptValue.toChatMessages(),
valuesForLLM,
callbackManager,
);
logger.debug('[FunctionsAgent] plan message', message);
return parseOutput(message);
}
}
module.exports = FunctionsAgent;