Видео

How does the model know which function to run? I have to explain to it somehow what a function does..

very cool

Hey, can we implement it with all together?

Does the structured output work with llm calls using bedrock?

but better if you can store several version of perspective and has calculation the benefit do a perspective than other perspective. because...in academic world there is several perspective to solve a problem. what you build here is only enhancement a perspective. event that we all here should appreciate this is a BIG STEP forward of improvement in field AI Knowledge. cheers...

The best 💯

Mistral will be big one!

thank you for the wonderful insights in the latest RAG developments. Can someone explain in simple terms the benefit of implementing "LangGraph" ? from what I understand it allows for more accurate LLM executions by limiting the "routes" the output of a certain LLM flows trough, improving it's reliability in execution. But why can't we empower LangChain "Agents" with the same functionality ? wouldn't the ideal agent have LangGraph capibilities built in ?

What happens if the graph gets stuck in a loop? (Web Search > not use full > Web Search > not usefull > ...) Do i have to add a "tries" counter to my state and end after x tries to prevent an infinte loop?

🎯 Key points for quick navigation: 00:00 *- Advance RAG control flow with Mistral and LangChain* 00:12 *- Combining small steps into comprehensive control flow for large language model applications* 00:25 *- Flow engineering uses a flow diagram to check response intent and construct answer iteratively* 01:06 *- Corrective RAG uses retrieval evaluator to assess document quality and trigger web search for additional information* 02:14 *- Hallucination note checks answer support by document, and answer question node checks generated answer relevance* 10:31 *- Bind MRAW to schema* 10:43 *- Convert JSON output* 10:59 *- Mock retrieval example* 11:12 *- Grading documents relevance* 11:25 *- Confirm binary score* 11:39 *- Define RAG chain* 12:05 *- Graph State explained* 21:11 *Adversarial Tax Routing* 21:52 *Hallucination Grader Defined* 22:18 *Router Conditional Edge* 22:47 *Web Search Fallback* 24:03 *Control Flow Implemented* Made with HARPA AI

what other types of functions? is there any good documention?

Mistral rocks!

In the last part, when the flow went twice to the web search tool, it basically searched on the same query, then how did it produce valid result 2nd time and not first time. How to ensure that it does not get stuck in the loop, because basically it does the same thing again and again without changing anything hoping to get correct result.

Thank you Sophia and Lance!

How does a request gets translated into llm input? Are you using special tokens to denote function call or response messages. Thanks for the help.

Thanks for this ! Learned a ton of good stuff, very well explained, will definitely be playing with your notebooks 😊 You’re fantastic for sharing such high quality work

nice

Calling the API costs money, right?

Awesome job! Thank you for sharing! What's the best way to do the RAG based on the relational database? We need to understand the question, go to the correct table of a database and find the most relevant records. Looks like we should support both keyword search and sematic search. For the keyword search, we need to extract the parameters, like the keyword, date of that question, the person who generated that record, etc.

Glad to see you in Mistral AI! 🥰

Excited to bring Mistral into Taskade with our upcoming Multi-Agent update! 😊

Can't wait to incorporate Mistral into Taskade in our next Multi-Agent update :)

What about this document be csv file? How can I do it?

What is the purpose of Mistral client? can we replace with a model run locally

You guys are amazing ❤❤

This is an amazing tutorial. so much valuable information packed in 30 min. Subscribed, Thank you!

Smokin hott

i don't want use it as a open source llm but instead i want it as local and deployed in my cloud service. I need to deploy it in the Azure cloud then what is the cpu and gpu requirement ???and can i use langchain.???

Guys, this is crazy good! Please don't stop your demos and explaining of concepts. If you read this - can you explain a lil bit more the concepts of action tools (usage, own implementations and so on). Thx in advance!

There's a mention of tool execution on the "model" side? What's the use-case for that?

I tried to do this with the OpenAI client and base_url set to my local Mistral-7b endpoint, basically using Mistral7b as a stand-in replacement for the OpenAI models. The tools format should be the same, right? It works with the gpt models but not with Mistral. Any idea, why?

Is function calling and system prompt compatible features? Setting tool_choice in "auto" but with usecase demanding a function call, the model write the JSON to call the function but includes it as a part of the content, instead of using tool calls explicitly.

*The Mistral AI impressed me a lot because it gave me a good code to train an AI without a gpu and using the technique of dividing the data.txt dataset into mini-batches freeing the memory at each training of the mini-batch, which came to work but unfortunately with an error at the end of the training of the last mini-batch of index out of range in self and could not solve it at all, But I was very impressed by the fact that she gave me this good code with few requests for hits (about 3 only)...*

is there functions calling support for typescript and Nextjs or only possible with python ?

Does Mixtral 8x7b have function calling or just the Large API model?

cool

Great, is Mistral 7B capable of function calling ?

I am getting this error at step 10, both on the colab and in my local interpreter. Any clue? ValidationError: 1 validation error for ChatCompletionResponse choices.0.finish_reason Input should be 'stop', 'length', 'error' or 'tool_calls' [type=enum, input_value='tool_call', input_type=str]

Congratulations on the launch of the channel ☺ great video, looking forward to the next ones!

Great ! can we change the ENDPOINT = "localhost" (or base_url) & api_key="NONE" ? It would be excellent !