Meta AI Llama 2 Chat 7B LLM LangChain Text Summarization Named Entity Recognition Colab Tutorial 🔥

#llama2 #metaai
Learn how to use Llama 2 Chat 7B LLM with langchain to perform tasks like text summarization and named entity recognition using Google Collab notebook.
Meta AI just introduced Llama 2. Llama 2 is available for free for research and commercial use.This release includes model weights and starting code for pretrained and fine-tuned Llama language models — ranging from 7B to 70B parameters. Llama 2 was pretrained on publicly available online data sources. The fine-tuned model, Llama-2-chat, leverages publicly available instruction datasets and over 1 million human annotations
If you like such content please subscribe to the channel here:

Relevant Links:

[skghtkt]

Hi Rithesh, thank you very much for the awesome video. I am a masters student and I will be starting my thesis soon . my topic is aspect based sentiment analysis on social media data using llm. I need your consultation. Is there any way to reach you ? Please let me know.

[luispodesta]

Thank you very much for your work. Is there any way to adapt it to use Call-2-7B-Chat-GGML?

[saitej]

Please share notebook and share steps how we can use LLAMA2 for fine-tune for NLP tasks : ner, keywords, qa, summarisation, embeddings

[choichoi]

Thank you so much for the wonderful video. I have a task where I have to input about 8000 worded prompts and have the model classify based on the prompt. Given the limitation of the llama2 input, would you have any recommendations for this problem? I am considering using Long-Llama2 or Llama2-7b-32K but wanted to hear your inputs!

[takuyayzu]

Thank you very much for the informative video! I have some questions if you don't mind:
1. How long did it take to get access to llama-2 on HF?
2. I've tried changing my runtime to use GPU hardware acceleration but I'm unable to see the amount of VRAM available. Is it going to display whenever I start running stuff in the notebook?

[narendrasingh-tgmb]

Hi Rithesh can we do few shot prompting and do relation extraction between NER entities

[djvxjzy]

Hello been watching your videos for quite a few weeks very helpful and insightful
I have a query regarding the NER part why we are getting the entity which are been related to a part person
I mean if a person name is Steve and his info and a list of random names and phone number or email I'd but only Steve relative data is been extracted and not the other entities how can I get those entities as well,
Also can you provide more templates so that we can experiment with more NER types.

[mdmonsurali]

@Rithesh
Since the release of LLAMA v2 I have been looking around for anyone sharing how to finetune it for german NER but I still haven't found any good tutorial. Could you please make a tutorial about it?

[steepehare]

Thank you for the video Rithesh. How well and accurately does this AI answer questions about a large text (book, novel)?

[thamilarasan]

How can we inject our own entities. And also fine tuning with our own dataset for ner. Please make a video for FT llama2 for ner

[kaoutarlakdim]

great video please I want to know if there is any way I can get the source PDF from which it took the answer along with the output

[satishmaddula]

the video is nice, i have a small question

what is the difference between and


can i access insted of ?

[rrqzusm]

Can I do the Name Entity Recognition for question answer pair custom dataset in the form of csv file using this approach?

[ayushvjain]

Hi Rithesh, I performed the steps the same way that you have shown in the video. But I am facing this issue where everytime I run the "print(llm_chain.run(text2))" my output of the entites is different and it tends to miss some entity here and there. How can I fix this?

[mehulparmar]

Which model from meta is best for generating code what this function/class is doing). Do you have any code sample also in python doing that?

[bnlnyqh]

How to limit the cuda usage while running locally.
Thank You

[janardhanb]

hi sir, i tried ggml model and using it via fast api, but when i send large text to the modeli get context size exceed error, please let me know how to overcome this, and is the max context size is only 2048?

[gidinated]

could you share the notebook please? thank you very much

[angelochu]

It seems to me this llama2 7b requires awful lot of vram!