Dify

Пікірлер

@nicomercurio77 3 күн бұрын

Hi, is the "Personalized Memory Assistant" model no longer available? I can't seem to find it. Thanks.

@HerroEverynyan 5 күн бұрын

Nice. This was exactly what I needed to get me started faster with the knowledge base feature. Thank you & looking forward to the next video as well!

@lysonober 6 күн бұрын

Nice 🥂

@user-hr3zd7oe3j 6 күн бұрын

👏🏻

@jtrtsay 6 күн бұрын

Heisenberg!

@meldonld 7 күн бұрын

Ohm

@anubisai 7 күн бұрын

Just what i was looking for. Thanks. 😊

@lindermannla 10 күн бұрын

Excelent content! Thnxz!

@betofoto 16 күн бұрын

Muito bom...

@user-hr3zd7oe3j 17 күн бұрын

🎉🎉🎉

@user-zu8bx7hq8k 17 күн бұрын

This is incredible!

@lysonober 20 күн бұрын

Nice

21 күн бұрын

Hey! :D I got it. Let's say the user asked something about an item (house, shoe, anything that varies stock (and sometime can be unnavalable). This dinamic content can't be on the knowledge. so I thought asking directly to the knowledge base. I created this simple api to send stock. 1st: I cant integrate, I'm doing smt wrong. 2nd: when trying to make a http request, passing the right values (Header n params) , I could-nt get the answer back in the output. I got status 200, but body unauthorized. Sorry if this is not the right channel, its just that you went right on the point :D great work, btw

@dify_ai 21 күн бұрын

Hello! I see what you mean here. Perhaps you need a code node to do some processing before you use the http request.

@dify_ai 21 күн бұрын

You can join our discord (link in description) and send some screenshot to show us what you are trying to do to better help you.

@Rodrigodemetrio Ай бұрын

omg excited to give it a try

@Riverhill-AI Ай бұрын

Great software! But why do users have to click on "start chat"? Why not just initiate the chat interface upfront? And, is the API compatible with OpenAI specs?

@user-hp9se2gx7e Ай бұрын

how i make one tool to save inputs for users?

@user-xf3kg4wf9e Ай бұрын

This link leads to an amazing resource! Read the full article (shortened: %SHORT_URL%)

@yichenwu6117 Ай бұрын

快加入graphRAG吧求求了

@GlobalUpdateNow37 2 ай бұрын

Great video. I just discovered dify. I am still learning it. Thank you for your tutorial. I am trying to use it with AItable and n8n

@DebasmitaMukherjee-i3q 2 ай бұрын

the "workflow process" bar is displayed to the user as well. how do i remove that?

@ademargr 2 ай бұрын

Hi! Me here again. First of all, I would like to know if there is any channel that I can question, such as GitHub or blog itself? My question today is: is it possible to do via API the inclusion of conditionals in the workflow and inclusion of new knowledge bases? As I said in the previous comment: I would like to attend several restaurants and they have different menus. I want it to be automatic for the restaurant owner to include the menu and I, as a developer, create a backend that includes, edit the workflow, etc. Can I?

@dify_ai 2 ай бұрын

You can join our Discord via link at the bottom of our official website or in the video descriptions. Feel free to discuss questions there.

@ademargr 2 ай бұрын

Is it possible at **Dify** I create a unique agent to serve customers of different restaurants, where each restaurant has a different menu, and according to a predefined trigger it searches on a specific basis?

@dify_ai 2 ай бұрын

Definitely possible, you can use a question classifier or If/else node to direct across different menus. You might need to define your own trigger though.

@Mouayad-dn6sy 2 ай бұрын

I have a lot of data like 5 GB on my computer and I don't want to download it to the site, can I access the data and use it without downloading it to the site?

@dify_ai 2 ай бұрын

You can try our community edition and deploy your own at github.com/langgenius/dify.

@mikew2883 2 ай бұрын

Very nice! Are there plans to add workflow integrations like onMessage Received in the near future similar to what VectorShift offfers?

@dify_ai 2 ай бұрын

Are you referring to an event trigger upon receiving a message? That's definitely in our roadmap! However, we probably don't have time to work on it for now as we're currently focusing on enhancing our multimodal capabilities. We welcome any contributions on GitHub.

@ademargr 2 ай бұрын

Sensational! I became a fan, let's evolve Dify!!

@3a146 3 ай бұрын

cute cartoon style

@jackbauer322 3 ай бұрын

is there a way to have a docker of the frontend shown her and deploy it elsewhere ?

@jackbauer322 3 ай бұрын

or even have the code for the frontend ?

@dify_ai 3 ай бұрын

@@jackbauer322 We are open source at github.com/langgenius/dify.

@jackgaleras 3 ай бұрын

Son grandes .. muy buen trabajo .. lo estoy probando , falta colocar como mas plantillas de flujos de trabajo .. ejemplos o algo asi ...

@jairocesaralexander5389 3 ай бұрын

Gracias por el video toma tu like

@jairocesaralexander5389 3 ай бұрын

ya lo probé eta muy bueno .. a este paso que sigue ??

@tonygan1132 3 ай бұрын

How does the Knowledge Retrieval part work? I guess I need to build a dataset for retrieving?

@dify_ai 3 ай бұрын

LLM can retrieve knowledge from dataset you uploaded to the plaform (Cloud edition). Check our documentation for more detail please docs.dify.ai/features/datasets.

@tonygan1132 3 ай бұрын

@@dify_ai Thx

@aianish 4 ай бұрын

Is this open source?

@ricardowillamys 4 ай бұрын

Parabéns!!! 🎉 Dify é uma ferramenta maravilhosa e está ficando cada vez mais completa. Agradeco aos desenvolvedores.

@poojainternational5360 4 ай бұрын

Hi, can you help build a tutorial to help deploy the Dify apps with user access control. The apps are currently are sharable with everyone and everyone can access the workflows.

@AI-OneNews 4 ай бұрын

I am waiting for the feature knowledge from url(websites etc.) that will take it to the next level!🎉

@nyacumo 4 ай бұрын

Hi, awesome product, a quick questin, im deploying with docker compose on a remote server, whenever i update on newer dify versions i lost changes or tweaks i made on. how could i keep those tweaks.? thanks in advance

@dify_ai 4 ай бұрын

Hello! You can check the Update Guide section in our release note at github.com/langgenius/dify/releases/tag/0.6.3. Don't wait to ask if you have any more questions.

@fleecejohnson4318 4 ай бұрын

fake apache 2.0 licensed project (due to the added verbiage) - This is not open source software so you should stop pretending it is. You should change to a proprietary license and stop deceiving people.

@TweedBeetle 4 ай бұрын

Could you help me understand what the changes mean?

@fleecejohnson4318 4 ай бұрын

@@TweedBeetle copy and paste the added verbiage into ChatGPT and ask it to give you a legal response on what it means.

@TheReconditeAlchemist 4 ай бұрын

Sorry to be the one to tell you but you’re not correct (from a code and development perspective). If you can see their source code, they are open source. The “fake” part isn’t them, it’s you. If you are trying to just repackage what they worked hard at building, launching their own product to compete against them… you are the only “fake” part of the equation. Learn to code, build something better, then open source it MIT… otherwise take a seat and ponder on the quote “it’s better to remain silent and thought a fool, than to speak up and remove all doubt”.

@jackgaleras 3 ай бұрын

creo que solo es un modelo de negocio

@antoniobfmoraes 4 ай бұрын

Beautiful work!

@JhonnyFdoLopez 5 ай бұрын

🎯 Key Takeaways for quick navigation: 00:00 *🧠 Uso de la arquitectura RAG en Dify* - La arquitectura RAG (Retrieval Augmented Generation) mejora el rendimiento de la aplicación de asistente de IA al equipar al modelo de lenguaje grande con conocimiento externo que le falta. 00:28 *📝 Configuración y procesamiento de texto* - La configuración de segmentación de documentos determina cómo se dividen los documentos en párrafos distintos, asegurando que las consultas de los usuarios puedan coincidir con los segmentos de texto relevantes. - Los ajustes de recuperación en modo de alta calidad permiten elegir entre búsqueda vectorial, búsqueda de texto completo y búsqueda híbrida, siendo esta última la mejor solución al combinar las ventajas de la búsqueda vectorial y de texto completo. 01:50 *🔍 Procesamiento de archivos y acceso a conocimiento externo* - Después de importar el conocimiento en la aplicación, el asistente de IA o generador de texto tendrá acceso al conocimiento externo y lo utilizará como contexto. - Se pueden elegir entre dos modos de recuperación: de uno a uno y de multipath, siendo este último más adecuado para aplicaciones con múltiples bases de conocimiento. Made with HARPA AI

@highmanbong 6 ай бұрын

Love it❤

@mayorc 6 ай бұрын

I do not see any template apps in Dify discovery, why? This happens both in the cloud version or the deployed (docker) version.

@dify_ai 6 ай бұрын

If you are looking for our templates apps, you can find them under the 'Explore' section in our cloud version at cloud.dify.ai/explore/apps. It is located at the top of the page. If you have more questions or need further assistance, feel free to raise them in our Discord channel discord.gg/PkWWYcRBbV.

@duynguyenkhac217 8 ай бұрын

Hello Dify i wonder that is there any way to use my own LLM like as a API ?

@dify_ai 7 ай бұрын

Of cource you can.Dify allows you to integrate any open-source or local models. What's even cooler is you can choose to deploy Dify locally.doc:docs.dify.ai/tutorials/model-configuration

@gericktoro 9 ай бұрын

Any chance to use a database instead of a text file as input?. Because it is a manual task, every time I need to update 15:49

@dify_ai 9 ай бұрын

Yes. You can now access your database as model input through the API, refer to the doc docs.dify.ai/advanced/extension/api_based_extension/external_data_tool

@UmairAllahwala-kb4ro 10 ай бұрын

Can we sell dify apps with free plan

@dify_ai 10 ай бұрын

You have to abide by our open source license when using it. And if you need to change brand logo or do SaaS multi-tenant mode, you'll need our commercial license. BTW, our license agreement has more detailed info. github.com/langgenius/dify/blob/main/LICENSE