- 827
- 840 846
Data Science in your pocket
India
Приєднався 11 чер 2019
Welcome to Data Science In Your Pocket! Dive deep into advanced AI and Data Science topics not thoroughly covered elsewhere. Discover comprehensive insights and answers to complex questions, all in one place. Subscribe for in-depth tutorials and expert knowledge on cutting-edge data science and AI.
We have tutorials on Generative AI, Reinforcement Learning, NLP, Time Series, Graph Analytics and other major Data Science domains
We have tutorials on Generative AI, Reinforcement Learning, NLP, Time Series, Graph Analytics and other major Data Science domains
Generative AI Interview Questions: Fine-Tuning LLMs
This video comprises all questions asked in GenAI Interview on LLM Fine-Tuning #ai #generativeai #interviewquestions #job #datascientist #ml
PDF Link : datasciencepocket.gumroad.com/l/xuisns
PDF Link : datasciencepocket.gumroad.com/l/xuisns
Переглядів: 33
Відео
How to extend Laptop RAM to run LLMs ?
Переглядів 1732 години тому
This video explains how to extend RAM in your laptop to run high end applications and AI models especially LLMs #ai #ml #datascience #cpu #gpu
NVIDIA Hymba : The best small LLM
Переглядів 1214 години тому
NVIDIA Hymba is a Hybrid Small sized LLM which has outperformed other small LLMs including Llama3.2 and SmolLM. It uses Mamba and Transformer architecture (hybrid) #ai #generativeai #llm #nvidia #ml
SAMURAI : Enhanced SAM2 for Object Tracking
Переглядів 6404 години тому
SAMURAI is a zero shot Object Tracking model, based on Meta's SAM2 which gives best results even in crowded spaces, Occlusion and Fast moving object. Checkout its details in this post #computervision #ai #ml #segmentation #datascience #deeplearning
Fine-tuning Multi modal LLMs (Llama 3.2 Vision)
Переглядів 2394 години тому
This video demonstrates how to Fine Tune Llama3.2 Vision (multi-modal LLM) using unsloth in Google Colab with code explanation #ai #generativeai #llm #ml #chatgpt
Which Multi-AI Agent framework is Best? Comparing AutoGen, LangGraph, CrewAI and others
Переглядів 5249 годин тому
This video compares top Multi AI Agent frameworks including AutoGen, Microsoft Magentic-One, OpenAI Swarm and others and explains which framework you should choose for your task #ai #openai #microsoft #chatgpt #framework #ml
TSMamba : Mamba based Time Series forecasting model
Переглядів 28314 годин тому
This video explains TSMamba, a Generative AI based Time Series forecasting model based Mamba architecture #ai #timeseries #generativeai #chatgpt #deeplearning #neuralnetworks
Microsoft TinyTroupe : New Multi AI agent framework
Переглядів 1,1 тис.16 годин тому
This video demonstrates Microsoft TinyTroupe, a new Multi-AI Agent framework comparative to LangGraph, CrewAI, AutoGen, OpenAI Swarm with example #ai #microsoft #openai #aiagents #autogen
Generative AI Interview Questions PDF
Переглядів 21516 годин тому
Generative AI Interview Questions Pdf with 160 questions and answers for topics like LLMs, Fine,-Tuning, RAG framework, Prompt Engineering, etc. Link : datasciencepocket.gumroad.com/l/xuisns?layout=profile #ai #interview #generativeai #ml #datascience
How to use Google Gemini-Exp-1114 for free?
Переглядів 1,2 тис.19 годин тому
This video describes how to use the latest best LLM, Google Gemini Exp 1114 for free #ai #google #chatgpt #openai #generativeai
Google Gemini-Exp 1114, The new best LLM
Переглядів 1,2 тис.21 годину тому
Google recently released Gemini-Exp-1114 which takes the top spot at LMArena beating GPT-4o. Check out its details and how to use it for free #ai #generativeai #llm #chatgpt #google
Generate Indian accent AI audio for free using python
Переглядів 10721 годину тому
Generate Indian Accent TTS audio for free using gTTS python library #ai #indianaccent #tts #chatgpt #googletranslate
How to use Qwen2.5 Coder ? The best coding LLM
Переглядів 917День тому
This video demonstrates how to use Qwen2.5 Coder using Ollama and python in the local system for free #ai #coding #programming #generativeai #python
Microsoft Magentic One: Multi AI Agent framework
Переглядів 1,4 тис.День тому
Microsoft Magentic One is a new Multi AI Agent framework which is simpler than AutoGen, LangGraph CrewAI and OpenAI Swarm #ai #microsoft #openai #chatgpt #autogen
Free AI tools platform : HuggingFace Spaces
Переглядів 13414 днів тому
This video demonstrates Huggingface spaces by Huggingface which provides many latest AI tools to use for free #ai #aitools #free #huggingface #generativeai
How to use GGUF LLM models using python?
Переглядів 22414 днів тому
How to use GGUF LLM models using python?
NVIDIA RAPIDS cuGraph : GPU acceleration for NetworkX, Graph Analytics
Переглядів 29614 днів тому
NVIDIA RAPIDS cuGraph : GPU acceleration for NetworkX, Graph Analytics
How to use Stable Diffusion for free ?
Переглядів 33621 день тому
How to use Stable Diffusion for free ?
Oasis : AI playable video game model
Переглядів 1,3 тис.21 день тому
Oasis : AI playable video game model
OpenAI SearchGPT: ChatGPT with Internet
Переглядів 10621 день тому
OpenAI SearchGPT: ChatGPT with Internet
Unlimited AI wallpaper generator for free using Stable Diffusion
Переглядів 31621 день тому
Unlimited AI wallpaper generator for free using Stable Diffusion
Manim : Python package for maths animation videos (3blue1brown)
Переглядів 84628 днів тому
Manim : Python package for maths animation videos (3blue1brown)
How is Claude3.5 enabled with Computer Use ?
Переглядів 138Місяць тому
How is Claude3.5 enabled with Computer Use ?
OpenAI Swarm : Multi AI Agent E-commerce system demo
Переглядів 1,1 тис.Місяць тому
OpenAI Swarm : Multi AI Agent E-commerce system demo
Run Flux Dev on Google Colab 8 GB GPU for free
Переглядів 1,7 тис.Місяць тому
Run Flux Dev on Google Colab 8 GB GPU for free
provide the note book
Hey can you please help me out with my project which is to detect number of leaves in a top view image of the plant in image processing im getting trouble to seperate the overlapped leaves 😢
Kindly drop an email at datasciencepocket@gmail.com
Wow this is super helpful!
Hey I want to use langgraph in my application but there is no clear official documentation, is there somewhere else I can refer to?
There is no docs unfortunately. Check this playlist instead: ua-cam.com/play/PLnH2pfPCPZsKbJwHRP0_l-rw5jllmBuuN.html&si=hdanlmmMpdOPXlin
@datascienceinyourpocket thank you
whats is link ?
arxiv.org/html/2411.11922v1
How to use this fine tuned model so we can have a fine tuned chat interface?
You can use gradio to build the interface. Just 1)push the model to hub 2) use locally in your repo. Whatever works for you
ple share the notebook bro
See pinned comment
Codes : colab.research.google.com/drive/1j0N4XTY1zXXy7mPAhOC1_gMYZ2F2EBlk?usp=sharing
Paper is out for more than 6 months, no Github link or model released. I wonder if the approach is stable as the paper is still not published yet. Let's see!
Yeah, I agree
I am using LangGraph. I find it most intuitive than AutoGen. But latest version of AutoGen I have not tried. What you suggest? Should I conintue using LangGraph or try AutoGen?
If you understand LangGraph, I guess you can leave AutoGen given the fact you know how to create custom tools
What app is this pls 😊
Character.ai
can I run my own trained model on this (model trained on flux-dev1-lora)?
Not sure, it has to be 8 bit quantized
share the colab notebook bruh
datasciencepocket.gumroad.com/l/jtzbtp
It looks promising. When is it planned to be released? Who will train and publish this model ?
I have a question, if I change the prompt I have to restart the piece of code that contains it every time. That piece of code takes a lot of time, which is added to the actual image generation time. I'm not familiar with Google Colab, but it seems to me that with each prompt change the wait becomes very long like this. Am I doing something wrong? Or is there a way to reduce the times? For example by moving a part of the code down, before print("Running denoising.")?
Nopes, you can't do it. This is actually a limitation of the code. But can't be helped out. If you understand --> as Flux Dev is huge and can't be fit in memory, we are loading its parts one at a time, using these parts and dumping them from memory. Unfortunately, the text-->embedding generator also gets dumped once used hence needs to be reloaded again when prompt is changed.
Great Video, thanks!
you got a new sub my friend and hope to see more videos soon
Thanks😊
Minimax ne génère pas en format tiktok, grosse erreur