Current location: Homepage > python > Generative AI Full Course – Gemini Pro, OpenAI, Llama, Langchain, Pinecone, Vector Databases & More - Ep194

Generative AI Full Course – Gemini Pro, OpenAI, Llama, Langchain, Pinecone, Vector Databases & More - Ep194

2025-07-11 11:26:48 [ai] Source: ByteGenius
I willopen my folder and intellipaat ai tutorialhere let me open myterminal so getclone and let's past the link and cloneit now I'll just go inside the folderend toendmedical chatbot using Lama 2 now I'minser myfolder now let's open my vs code here soif you have pyam or any other codeeditor you can use it feel free to useit noissue um guys can you confirm my vs codeis visible to all ofyou can you seethe text and all clearly you can let meknowokay so the first thing what I need todo I'll be creating one virtualenvironment okay so let's create onevirtualenvironmentuh yes yes uh we'll be doing on CPU okaythat iswhy and yesterday I told you to downloadone particular model guys it's around4GB I think youremember uh yeah see uh the same thingwe can do it on the neural lab as wellokay so you can also open up your neurallab and you can also do it there but thething is like there actually I need toupload my model and it will take timeokay it will take time to upload mymodel there so I will show you how to doit how to set up the environment here aswell so let me open my neurallab because it's around 4GB model so ifI want to like upload there so it willtake time so that's why I'm showing onon local machine so start my lab so fromhere actually what you can do you can uhlaunch up this onepython so this is my medical chatBo so here also you will get the sameenvironment as as your V code let meshow yousee allright okay but I have the model in mylocal machine I already downloaded butif I want to upload it it will take timeso it's better to use my uh this vscodeokay so now let me create theenvironment so just write the commandcondcreate hypen in um I'll just name theenvironment asuhmedical or I can just write M chatbotthat means medicalchatbot it's very EAS just open up yourterminal and just write code space dotokay this is the command to open the uhvscode okayyeah and now let's uh take the pythonversion equal to3.8 hyen y so everyone you should usePython version 3.8 okay no you don'tneed to use uh like less than 3.8otherwise you might face some issue okayyou can also take 3.9 it's fine but I'llbe using 3.8 this specific version alsojust let me mention the command in myrme file so here step torunsteps torun theproject so the first thing what I needto do I need to createone so I can just write thisline I just need to create myenvironment now let me just quicklycreateityou can use it okay you can use it it'sup to you what particular version youwill be using it's up to you personallyI like python 3.8 because it supportslikeBuh guys am I audiblenowokay uh sorry there was a small powercut from my side extremely sorry forthat okay thankyou uh yes uhokay see uh first of all you need tocreate one environment okay so this isthe command you need to execute uh thisis the command you need to execute justConta create hypen in then your uh nameof the environment then use Python 3 uh3.8 okay and then create the environmentthen I just need to activate theenvironment okay so this is the commandso justcopy and pasteit now it has been activated so let mealso uh give the commandhereokay so environment creation is done nowI need to install the requirements okaysome of the requirements I need for thisproject so let's install therequirements so here I'll be creatingone file calledrequirement.txt okay and now let's mention uh therequirements so the first requirement Ineedhere uh something called C transformforer I'll tell you why this CTransformer isrequired um so the first thing I needsomething called C Transformer okay andI will be using this specific version ofthis C Transformer so you can uh use anyof the version but I'll be using thisparticular version because I also wantto show you like whenever you arecreating any kinds of projects okay youalso need to specify the version of thelibrary you are using let's say thisproject you are sharing after let's sayone year okay so what will happen at theTimes uh some Library changes would behappen Okay and some of thefunctionality would be removed some ofthe functionality would be replicated soit's better to use a specific versionalways okay so that is why you can usespecific version let's say if I searchthe C Transformer on Google CTransformerPi so you will see differentdifferentuh version of the C Transformer Libraryso release story now see guys differentdifferent and this is the current one0.227C Transformers that means see the modelactually I'm going to use it's aquantized model because we'll be runningon CPU so that is why uh we need this CTransformer library to load thequantized model gotit I think yesterday you saw like we areusing something called Lama CPP libraryright but here we are using Lang chainand if you want to use Lang chain so youneed to use the C TransformerLibrarythen the second Library I need sentenceTransformer okay because I want todownload this uh model from the huggingfaceitself uh uh like which model I'll bedownloading from huging face itselfbecause here I need one embedding modelbecause as the architecture I showed youhere we'll be generating embedding okaywe'll be generating embedding of ourtext SS and to generate this embeddingwe need one embedding model okay sowe'll be using one free embedding modeluh guys just a minute just aminuteokay uh now I think my network isfineokaygreat okay fine so that is why actuallyuh for this embedding okay for thisembedding generation actually I need oneembedding model okay and that particularmodel I'll be downloading from the hface itself okay so that is why I needthis seat uh Transformer sorry sentenceTransformer Library gotitthankyou now let's uh see our thirdrequirement actually I'll beusinguh third actually I need to use uh thispine cone client because I want tointegrate pine cone database okay Pinecon Vector DB so that's why this pinecone client is needed then as I told youI'll be also

(Editor: go)

Recommended articles
Hot reading