Flowise: How to chat with your PDF using no-code UI Framework
ฝัง
- เผยแพร่เมื่อ 2 มิ.ย. 2023
- [NEW] • Flowise/Langflow: How ... demonstrate the code and logic behind this Flowise chatbot. Please feel free to check it out.
Thank you for watching. Hope you have a nice day.
A Tutorial for Flowise UI
Flowise is a TS/JS no-code UI tool for LangChain , if you are familiar with JS, please feel free to contribute to this project.
github.com/FlowiseAI/Flowise
About me:
Welcome to my TH-cam channel! As an AI engineer working at Tech Giant, I am excited to share my knowledge and insights with you. Join me as we delve into the fascinating world of artificial intelligence, machine learning, and everything related to cutting-edge technology.
On this channel, you can expect in-depth tutorials, demonstrations, and discussions about various AI topics. From overviewing popular AI packages and frameworks to diving into the latest research papers, I aim to provide valuable content that helps you stay updated and empowered in the fast-paced AI industry.
Make sure to hit the subscribe button so you won't miss any of my upcoming videos.
PS: I am new to TH-cam, and still improving my video editing skills. 😊
Thank you, we need more videos like this to understand the possibilities
Thank you so much for watching. I am new to TH-cam and still improving my video quality.
Great video, thanks
Awesome!!! Insta subscribed! Thank you very much for sharing, I will contribute with you.
Thank you for watching!!!
great! keep sharing with us your knowledge and skills
Thank you so much.
Круто, 13:00 жду с нетерпением :)
Best yet! Thank you!
If the pdf has a structured table, will the split and encode destroy easy extraction of meaningful relationships in a query?
Do you think i can setup flowise to read a folder locally for csvs looking for a name and then using those csvs or pdfs i guess, to talk about data. Instead of the built in download for csv agent.
How do you add a prompt to the LLM chain so it can answer from documents?
Timestamp
1. Build a basic Chatbot: th-cam.com/video/ROoGbFrvuwM/w-d-xo.html
2. Build a Chatbot to Chat with PDF: th-cam.com/video/ROoGbFrvuwM/w-d-xo.html
How can i use Chat Prompt Template on ChatOpenAI with memory?
After replicating your structure I am getting 0 in the response of all my prompts
Where am i going wrong? Please explain
Two questions
1 - Why do you prefer Hugging Face Inference for Embeddings?
2 - Do not need to put the model for embedding?
Hi Thank you for your question.
Q1: Hugging Face Embedding is free, while OpenAI Embedding will charge fee based on the number of tokens. If we embedding PDF files with few hundreds of pages, the fee might be out of expectation.
Q2: We do not need to put the model for embedding, as there is DEFAULT_MODEL_NAME. You can check Flowise source code here github.com/FlowiseAI/Flowise/blob/ff93d11913e00ef2daa390757a2a0ed14485c234/packages/components/nodes/embeddings/HuggingFaceInferenceEmbedding/HuggingFaceInferenceEmbedding.ts#L3, it use the HuggingFaceInferenceEmbeddings, which has the DEFAULT_MODEL_NAME, please refer here python.langchain.com/en/latest/_modules/langchain/embeddings/huggingface.html (python code and LangchainJS should be the same logic).
@@techwithray8943 Thank you
Great content. question! I tried it and it seems answers are restricted to the content of the doc and can’t go out of context. Is there any way around that so it can respond like the normal ChatGPT and also refer to the doc?
Yes, you are right!! It cannot go beyond the content you have. Cuz the purpose of this is to find content in you PDF, not from ChatGPT. If you really want to go beyond your PDF, one solution: you can wrap your search function to Function Calling from ChatGPT endpoint, you can get some ideas from th-cam.com/video/gtbj4AgOkTo/w-d-xo.html. Hope this is helpful for you.
Hello, 🎉 great content! I would like to request a video on a solution for chatting on PDFs using an open-source model from Hugging Face that utilizes GPU and ensures data privacy.
Thanks for watching. I will work on this request and record another video.
Have you successfully debugged with azure openai api? I tried several times, it was fine with openai but not successful with azure
Hi, I did not use azure openai api. It should be simple to use it. Let me try it first and get back to you. Thanks.
@@techwithray8943 Thank you, but I just succeeded after changing a set of APIs. Obviously, there is a certain difference between using azure and native openai to generate content😄
@@liuchuchenliu2802 I am so glad that you figured out. I checked the azure oai services, the authentication is a bit different and the gpt version might be different from what we choose from openai.
good AI
Thank you!!
perfect for searching Turing..can i search other words besides Turing?
Yes, you can search another words, it will give you other related content. You can check this video, it shows more example th-cam.com/video/MERQHJBCq4c/w-d-xo.html.
how to combine Prompt and PDF
Hi thanks for watching. You can use LLM Chain to directly use prompt, or use Conversational Retrieval QA chain to add "inputs" along with embedings and send to OpenAI API.