Thank you - I work for a pharmaceutical company - your videos are very useful - and even though I have a medical background - zero IT - I may follow your instructions without problems.
Can you make a video in regards to using Large language models for SQL query? For example i'd like to see how to use a LLM to produce text to sql to query a database, like a mysql database. You can ask it questions like "how many users went offline on Monday?" it will then translate that to sql query, query the database, return the data and then return the data to the user in a natural language. Please if you can make a guide on this bro! I love your channel, you have earned my sub!
hello? I am a professor at a medical school, and I find your lectures interesting. 1. If I change the PDF file in GitHub's 'data' foder to a PDF file for a disease in another category, will it become a RAG app for that field? 2. How much difference will there be in performance between using the openAI GPT4 model for the PDF file and using the Meditron 7B model? 3. I am asking this question because my knowledge is limited, but rather than just embedding the PDF file in the 'data' folder in LLaMA2 and using it as a chatbot, I am developing a chatbot using meditron 7B LLM trained for medical purposes. Is it more effective?
Very good video. Quiet detailed. Following your instructions, i might take a try into optimizing a RAG. No idea about coding nuch, but as doctor, i will give it a go
I want an open source model for generating diagnosis plans based on user input. I have dental data can you suggest any open source model. The don't want to use cloud as data is sensitive
You can fine-tune an LLM for that task which will have verbiage derived from your dental data.... In case of generating diagnosis from a set of data/docs, you can also implement a RAG system.
A great option for a local, offline AI/LLM is the use of Ollama software to run these models on your computer without sending anything to the Internet or third-parties.
Great tutorial. I learn alot from your videos. I want to select which llm & embeddings will be best for education. It is required to generate assessments with solutions out of learning material to help teachers mentor students.
I really appreciate your tutorial and I am big fan of your videos....is it possible to give some example where using RAG as service - using FAST API or Flask, and how to use offline models - since most of the places we dont have access to external internet from organizations and need to implement everything as service - based architecture.. it will be greatly helpful..
@@AIAnytimeThanks bro:) it’s just i am noob at this, so it’s not obvious for me what to do, although i am practicing oncological surgery , and doing phd in oncology so I’m very much interested in the project. Big ups! 😊 Edit I suppose it’s also possible with llama.cpp right ?
Thank you - I work for a pharmaceutical company - your videos are very useful - and even though I have a medical background - zero IT - I may follow your instructions without problems.
It's great to develop a RAG service in the medical field. I am also very inspired by your attempts. Thank you for always posting good content.
Thank you so much 🙏
Can you make a video in regards to using Large language models for SQL query? For example i'd like to see how to use a LLM to produce text to sql to query a database, like a mysql database. You can ask it questions like "how many users went offline on Monday?" it will then translate that to sql query, query the database, return the data and then return the data to the user in a natural language. Please if you can make a guide on this bro! I love your channel, you have earned my sub!
hello? I am a professor at a medical school, and I find your lectures interesting.
1. If I change the PDF file in GitHub's 'data' foder to a PDF file for a disease in another category, will it become a RAG app for that field?
2. How much difference will there be in performance between using the openAI GPT4 model for the PDF file and using the Meditron 7B model?
3. I am asking this question because my knowledge is limited, but rather than just embedding the PDF file in the 'data' folder in LLaMA2 and using it as a chatbot, I am developing a chatbot using meditron 7B LLM trained for medical purposes. Is it more effective?
Very good video. Quiet detailed. Following your instructions, i might take a try into optimizing a RAG. No idea about coding nuch, but as doctor, i will give it a go
Sounds great...
I want an open source model for generating diagnosis plans based on user input. I have dental data can you suggest any open source model. The don't want to use cloud as data is sensitive
You can fine-tune an LLM for that task which will have verbiage derived from your dental data.... In case of generating diagnosis from a set of data/docs, you can also implement a RAG system.
A great option for a local, offline AI/LLM is the use of Ollama software to run these models on your computer without sending anything to the Internet or third-parties.
Best guy on TH-cam for Generative AI
Thanks broda
Great tutorial. I learn alot from your videos.
I want to select which llm & embeddings will be best for education.
It is required to generate assessments with solutions out of learning material to help teachers mentor students.
I really appreciate your tutorial and I am big fan of your videos....is it possible to give some example where using RAG as service - using FAST API or Flask, and how to use offline models - since most of the places we dont have access to external internet from organizations and need to implement everything as service - based architecture.. it will be greatly helpful..
Check my RAG playlists. There are tons of such videos. This is a RAG where we have used FastAPI with an offline model.
Hello, nice tutorial. How did you download the meditron model to local?
Great video! .......... i am not able to down load llm model and embedding model from provided link how do i do it
Great video, how we can run the code using Nvidia DGX A100 machine?
Great Info Sonu!!! You are the best!!!!
Thanks a ton
Thanks for awesome vid and tutorial as usual! How can I run this using GPU, and not CPU? As the script doesn't work with M2 Mac..?
Metal? Cublas support.... CTransformers support that. Check the GitHub.
@@AIAnytimeThanks bro:) it’s just i am noob at this, so it’s not obvious for me what to do, although i am practicing oncological surgery , and doing phd in oncology so I’m very much interested in the project. Big ups! 😊
Edit I suppose it’s also possible with llama.cpp right ?
can we run two LLM models in one script?? like Mistral 7B and this one at a time?
Absolutely.
Keep these coming
Is there any equivalent model for the legal domain?
LegalBert? Otherwise you can use general purpose LLMs.
good video
Glad you enjoyed
Wow ❤
Thanks
thanks
No problem