Hey David, would it be possible for you to make the code available again? It would be greatly appreciated, as it was a valuable resource for understanding the implementation in more detail. BTW thanks for your excellent content and for sharing your knowledge with the community!
Thank you so much for this. are you going to cover for other benchmarks?, or same methods could be applied as well ?. Also, learned something new, setting a cache_dir path so you can know where they're located and when not needed you can delete it to save space. Thanks.
The ChatOpenAI function accept models other than openAI ? i saw this from langchain_community.llms.mlx_pipeline import MLXPipeline llm = MLXPipeline.from_model_id( "mlx-community/quantized-gemma-2b-it", pipeline_kwargs={"max_tokens": 10, "temp": 0.1}, ) Thanks
Any call to openai api will cost tokens. Will langchain help to reduce the tokens being consumed? Or langchain only consolidates all the request but same token will be consumed compared to asking them 1 by 1?
Thanks for this video David. I created a python application which functions as a chatbot. Trained on some PDF-files and linked to openai to give users answers based on this pdf. For four days, i am a noob with regards to coding/hosting/etc., I have tried everything from Replit, google cloud, Heroku to deploy my application to the web. It just doesn't seem to work. could you maybe give me some tips on how to do this? Are you familiar with this?
It’d be a little complex but it’s possible. ChatGPT can return responses in a tabular format, so you could prompt engineer your way to this. The one challenge would be converting the response from OpenAI for appropriate use. Even though ChatGPT can return tabular responses, they are done in the form of markdown strings. So you’d have to convert the markdown strings into something else first. Here are the precise steps I’d take: 1. Load in the PDF file with your data. 2. Generate a prompt asking ChatGPT to extract information and return in a tabular format. 3. Figure out a way in Python to take that tabular string and convert it into a Pandas dataframe. 4. Save out the Pandas dataframe as an excel file using the built in “to_excel()” function. Just be aware that if you do this, you are giving your own proprietary data to OpenAI who stores it for 30 days.
Hey David, great video! I have an idea of creating a Project Proposal Generator where a User specify basic things like Objectives and Goal etc, and it creates the Roles/Responsibilities etc in the Project Proposal. Would love to see if you can make it!
Hi there Jack. It’d be totally possible to make that as it’d use basically the same framework that I used for the latter project. I don’t have the time to code it up myself, but the code that built the Business Profiler is linked in the description of this video, and I also just did a live stream on May 5 walking folks through every step for enabling something like this. 😃
Thanks a lot David. That was really helpful. AWS is evolving super fast and the deep dive Sagemaker videos from amazon are already super old now. Looking forward to the next videos of yours.
Missed the stream, but tuned in to the video. Informative video. I think you explained concepts well. This is coming from me - a beginner with no data science background. Happy birthday, by the way! Hope you have a great day today. Looking forward to more videos! - Kyan V.
U a legend for going through there work and explaining it turning it up side down mechanic style 😎
Hi, great video! Can you share the kaggle link?
@26:13 actually this is one the most interesting part of this stream
Hey David, would it be possible for you to make the code available again? It would be greatly appreciated, as it was a valuable resource for understanding the implementation in more detail. BTW thanks for your excellent content and for sharing your knowledge with the community!
You're good Man!
21:36
Good Detailed tutorial - well explained
Thank you so much for this. are you going to cover for other benchmarks?, or same methods could be applied as well ?. Also, learned something new, setting a cache_dir path so you can know where they're located and when not needed you can delete it to save space. Thanks.
ahhh, cache_dir not working.
The ChatOpenAI function accept models other than openAI ? i saw this from langchain_community.llms.mlx_pipeline import MLXPipeline llm = MLXPipeline.from_model_id( "mlx-community/quantized-gemma-2b-it", pipeline_kwargs={"max_tokens": 10, "temp": 0.1}, ) Thanks
Thanks for explaining everything so clearly and thoroughly
Thanks a lot! great session
How do you make a twitter cash tag?
Is it just me or is the video stretched vertically?
Not just you. I explain why at the outset. 😅
Thank you for expressing your thoughts and experiences. Much appreciated. Keep posting.
real bullshit! Binary & K-Means, no way😅
please give source code
Hi Sir where is the source code Share ir Please!
Sir where is the source code please share it
David, great video. Keep it going! I'd be great to walk through some code on the video btw.
Great video!
promo sm
Why would I follow you if you don't even allow pelple to save your videos?
I don't understand why it takes so long for the beginning of the video to wait
Such a good video! Thanks!
Any call to openai api will cost tokens. Will langchain help to reduce the tokens being consumed? Or langchain only consolidates all the request but same token will be consumed compared to asking them 1 by 1?
Great content btw
Someone would up with a way to where we can bypass all of that.
Can you do a flowise with Gradio for UI and deploy it for a real time use case?
Thanks David. Appreciate your insights and time spent on the video.
Thanks for this video David. I created a python application which functions as a chatbot. Trained on some PDF-files and linked to openai to give users answers based on this pdf. For four days, i am a noob with regards to coding/hosting/etc., I have tried everything from Replit, google cloud, Heroku to deploy my application to the web. It just doesn't seem to work. could you maybe give me some tips on how to do this? Are you familiar with this?
Hey, can you suggest any tutorial for chatbot application using pdf
How to extract particular data like names or contacts from pdf files and save in excel sheet using Langchain, Open AI and ChatGPT?
It’d be a little complex but it’s possible. ChatGPT can return responses in a tabular format, so you could prompt engineer your way to this. The one challenge would be converting the response from OpenAI for appropriate use. Even though ChatGPT can return tabular responses, they are done in the form of markdown strings. So you’d have to convert the markdown strings into something else first. Here are the precise steps I’d take: 1. Load in the PDF file with your data. 2. Generate a prompt asking ChatGPT to extract information and return in a tabular format. 3. Figure out a way in Python to take that tabular string and convert it into a Pandas dataframe. 4. Save out the Pandas dataframe as an excel file using the built in “to_excel()” function. Just be aware that if you do this, you are giving your own proprietary data to OpenAI who stores it for 30 days.
@@dkhundley Thanks for reply, Can you please make one video on this topic so that I can follow the mentioned steps.
Hey David, great video! I have an idea of creating a Project Proposal Generator where a User specify basic things like Objectives and Goal etc, and it creates the Roles/Responsibilities etc in the Project Proposal. Would love to see if you can make it!
Hi there Jack. It’d be totally possible to make that as it’d use basically the same framework that I used for the latter project. I don’t have the time to code it up myself, but the code that built the Business Profiler is linked in the description of this video, and I also just did a live stream on May 5 walking folks through every step for enabling something like this. 😃
@@dkhundley Awesome, could you share the link of that specific Livestream, would love to check, also thanks for your response.
About 54 min in your sound cuts out.
You doing a really great job on your channel! You should use a service such as *Promo sm*!
Thanks a lot David. That was really helpful. AWS is evolving super fast and the deep dive Sagemaker videos from amazon are already super old now. Looking forward to the next videos of yours.
A white rabbit ...
😇 Promo>SM!!
I’m also looking forward to that new Kirby game!
Skip forward a minute. I forgot my headphones weren't paired to my computer for that first second. Sorry about that. :)
Missed the stream, but tuned in to the video. Informative video. I think you explained concepts well. This is coming from me - a beginner with no data science background. Happy birthday, by the way! Hope you have a great day today. Looking forward to more videos! - Kyan V.
I missed the livestream but definitely glad this was available for me to watch!