0% found this document useful (0 votes)
13 views4 pages

Agents

Autonomous Language Agents

Uploaded by

Marcos Luis
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
0% found this document useful (0 votes)
13 views4 pages

Agents

Autonomous Language Agents

Uploaded by

Marcos Luis
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 4
9115124, 8:53 AM GitHub - aivaves-cr/agents: An Open-source Framework for Data-centc, Sal-evolving Autonomous Language Agents Bi aiwaves-cn/agents Pile [An Open-source Framework for Data-cenvic Self-evolving Autonomous Language Agents mB Apoch ty Sesto Y 40BVorks PF Branches ts © ga Acty 1 Neteations © Code © tssves 3811 Puleequests ©) Discussions © Actons fH] Projeas OD Securty lw Insights BF mater > | PABonches Qong PO Q Goto ie Gotofie ode @ Michsetzhowwang. Werge pul eavest #205 rom Olexandr@6/oatch-7 crea Saapraga © Oe assets Update README md 3 months ago f eamptes faint esto ron 3 months ago i seiagents faint es to uni 3 months ago Di enveramote faint fs to runt 3 months ago Di stignore faint fs to runit 3 months ag0 Di ucense ‘Create LICENSE last year Di Reames Update README 5 daysago Bi reauirementst Upload agents2o 3 months ago D setupey Upload agenis2o 3 months ago AGENTS Agents 2.0: Symbolic Learning Enables Self-Evolving Agents 2 Project [2 Paver) { Docs] {Overview} [Installation] [D> Citation CD RAOME 1 Apache 20K ANews ‘+ (2024.06.25) We release Agents 20, a major update tothe origina bray, adding suppor for agent learning and evaluation. htps:igthub.com/aivaves-cnlagents ue 9115724, 553 AM Chub - aiwaves-cvagens: An Oper-source Framework for Date cen, Sal evolvng Aulonamous Language Agents "Overview [Agent symbolic leaming isa systematic framework for taining language agents, which is inspited by the connectionst leaning procedure used {or taining neural nets. We make an analogy between language agents and nevral nets: the agent pipeline of an agent corresponds tothe ‘computational graph of a neural net, 2 node inthe agent pipeline corresponds to liye in the neural net, and the prompts and tools for 2 nd correspond tothe weights of slayer. In this way, we are abe to implament the main components of connectionst learning, ie, backward propagation and gradient-based weight update inthe context of agent training using language-based loss, gradients, and weights Computation Neural Netwok _Learnable Neural Nets Connectionist Learning Graph Layer Weights Numeric Optimizer Gradient & Loss ; y u u ee: [l E @ & ox tT ? t Agent Pipeline Node Prompts & Tools Symbolic Optimizer Language Agent Symbolic Learning Gradient & Loss \We implement loss function, back-propagation, and weight optimizer inthe contest of agent training with carefully designed prompt pipelines For training example, our framework fist conducts the “forward pass" (agent execution) and stores the input output, prompts, and tool usage in each node ina “trajectory” We then use a prompi-based loss function to evaluate the outcome, resulting ina “language loss ‘Afterward, we back-propagate the language loss from the last tothe first node along the trajectory, resulting in textual analyses and reflections for the symbolic components within each nade, we call them language gradients. Finally, we update all symbolic components in each node, a5. well a5 the computational raph consisting ofthe nodes and their connections, according tothe language gradients with anathercarefuly designed prompt. Our approach also naturally supports optimizing multiagent systems by considering nodes as diferent agents or allowing mukiple agents to take actions in one nade, Workflow Illustration htps:igthub.com/aivaves-cnlagents 2a 918124, 8:53 AM Installation Installation from git repo branch: Installation fo local development: ritps:igthnub.convaiwaves-cnfagents srawork for Date-centi, Selfevolving Autonomous Language Agents ‘9115124, 8:53 AM GitHub - aivaves-cr/agents: An Open-source Framework for Data-centc, Sal-evolving Autonomous Language Agents oh Star History ‘vt Star History aa Releases Packages Contibutors 24 CO °vOseeTOGCCHHs Languages htps:igthub.com/aivaves-cnlagents

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy