DataStreaming with LangChain & FastAPI

No video

DataStreaming with LangChain & FastAPI

Рет қаралды 17,214

Coding Crash Courses

Күн бұрын

Пікірлер: 38

@code.scourge 10 ай бұрын

Danke Bruder, habe dazu nichts bei der Langchain doku gecheckt.

@hopeok1225 4 ай бұрын

Love the video! How would you stop (interrupt) streaming with the FastAPI example and ensure that openAI stops generating tokens before it’s done?

@vignesh462 8 ай бұрын

Woov..You made it very simple. Thanks

@robertmazurowski5974 Жыл бұрын

Great tutorial, Thanks

@codingcrashcourses8533 Жыл бұрын

Thank you for your comment

@quadhd1121 6 ай бұрын

Now can use the backend with cloud run , lambda or some serverless function

@codingcrashcourses8533 6 ай бұрын

Is that a question? :)

@devtoro5601 29 күн бұрын

although it works, i think you didnt follow the [content-type=text/server-event] response format.

@codingcrashcourses8533 29 күн бұрын

Glad you found my new video on streaming^^

@phani3519 3 ай бұрын

do you have code for using a retrievalQAchain to send streaming responses?

@henkhbit5748 5 ай бұрын

Thanks, would be nice 2 see with opensource llm and streamlit OR gradio as the frontend.

@codingcrashcourses8533 5 ай бұрын

I use fastapi to not be forced to use Gradio or streamlit ;)

@riteshpanditi3635 Ай бұрын

Using astream, the response from the LLM has words that are split for example the word "hippopotamus" comes as 2 chunks "hippo" and "potamus". When creating an app, how to recognize and combine the 2 split parts into a single word for front-end?

@codingcrashcourses8533 Ай бұрын

does not matter as long as there is no "space" Token in between :). The word will just be concatenated

@waqasobeidy8318 Жыл бұрын

This was great. Thanks for sharing.

@codingcrashcourses8533 Жыл бұрын

glad you liked it :-)

@ali.shah.repository Жыл бұрын

Always enjoy your videos. Would love a deep dive on agents. Also is there a way to contact you regarding consulting?

@codingcrashcourses8533 Жыл бұрын

sorry, I dont do consulting

@emmanuelthiery2459 Жыл бұрын

I used langchain to make several app applications for customers, what do you need ?

@arslanabid2245 7 ай бұрын

How can we create FASTAPI for langchain's chat-with-pdf chatbot ? please help! I really like your method of teaching

@codingcrashcourses8533 7 ай бұрын

I dont have a Video on that. I Cover this more Real World topics currently only in my udemy course, since its more or less advanced and only experienced devs are interested in that

@arslanabid2245 7 ай бұрын

@@codingcrashcourses8533 please provide that udemy course link. i really want to create api for the chat-with-pdf project of mine

@thewimo8298 7 ай бұрын

The streaming is only by tokens and putting a space inbetween that, how do you remove the space between tokens but keep the space between words though?

@codingcrashcourses8533 7 ай бұрын

Sorry dont know what u mean. The tokenizer does that

@DanielWeikert Жыл бұрын

When using langchain llms Is it just a kind of wrapper for the transformer eg like """ from transformers import LlamaForCausalLM, LlamaTokenizer tokenizer = LlamaTokenizer.from_pretrained("/output/path") model = LlamaForCausalLM.from_pretrained("/output/path") """ Can you elaborate on that? My current understanding is that they refer to this for "not openai models" Thanks and br

@codingcrashcourses8533 Жыл бұрын

whats exactly your question? You should be able to use that, yes :-)

@aachalpatil290 Жыл бұрын

hi, can you please explain how can we do this while working with more than one input variable.

@codingcrashcourses8533 11 ай бұрын

The API or Langchain? In LangChain you just pass a dictionary {input1: val1, input2: val2}. In FastAPI you have multiples ways, I would suggest the use a Pydantic class with multiple attributes