- Видео 46
- Просмотров 123 853
Victor Dantas
Добавлен 3 сен 2011
I talk about Cloud, AI and stuff. Mainly from Google, because that's where I work.
This channel's manifesto:
- No clickbait titles
- No sensationalist thumbnails
- No content fluff
Buy me a coffee: buymeacoffee.com/vdantas
LinkedIn: LinkedIn.com/in/vicdan
This channel's manifesto:
- No clickbait titles
- No sensationalist thumbnails
- No content fluff
Buy me a coffee: buymeacoffee.com/vdantas
LinkedIn: LinkedIn.com/in/vicdan
Deepseek R1 vs Gemini 2.0 Flash Thinking (Reasoning models on Google Cloud)
Let's talk about how to deploy Deepseek R1 on Google Cloud, how much that would cost, and how it fares against Gemini 2.0 Flash Thinking.
Note that the price information I showed for running Deepseek R1 70b on Google Cloud considers a node running 24/7, in standard provisioning mode, without any discounts. You may be able to obtain a lower price. It is likely however that you need high volume usage to justify the cost of hosting the model yourself.
Some additional notes regarding Deepseek vs Gemini Flash thinking (forgot to mention in the video):
- Deepseek R1's context window is 128k tokens, max output size is 4k tokens
- Gemini Flash thinking's context window is 1M tokens, max output size ...
Note that the price information I showed for running Deepseek R1 70b on Google Cloud considers a node running 24/7, in standard provisioning mode, without any discounts. You may be able to obtain a lower price. It is likely however that you need high volume usage to justify the cost of hosting the model yourself.
Some additional notes regarding Deepseek vs Gemini Flash thinking (forgot to mention in the video):
- Deepseek R1's context window is 128k tokens, max output size is 4k tokens
- Gemini Flash thinking's context window is 1M tokens, max output size ...
Просмотров: 65
Видео
Gemini 2.0: announcements, demos, and stuff you need to know
Просмотров 635Месяц назад
A roundup of announcements around Gemini 2.0, including: - Gemini 2.0 Flash (native audio output, native image output, native tool use, multimodal live streaming) - How to access new models - Performance * Pricing - What is available today to everyone, to trusted testers, and what is coming soon - A quick overview of the new reasoning model Gemini 2.0 Flash Thinking (more on that in a future vi...
Gemini-exp-1206 and other Gemini Experimental Models (Available in Google AI Studio)
Просмотров 927Месяц назад
Just a quick look at gemini-exp-1206 and some other experimental models in Google AI Studio: what are they, how they work. I put gemini-exp-1206 to test with a coding example. Experimental models are free of charge (but can't be used in Production). Not the best audio quality in this one, sorry about that. Confused about Google AI Studio vs Vertex AI vs Gemini? Watch this: ruclips.net/video/MRD...
Prompt Optimization on Vertex AI (+ Langchain promptim)
Просмотров 5702 месяца назад
Still doing prompt engineering yourself? Let's talk about how having LLMs do it. Vertex AI Prompt Optimizer: cloud.google.com/vertex-ai/generative-ai/docs/learn/prompts/prompt-optimizer Langchain promptim: blog.langchain.dev/promptim/ #genai #llmops #vertexai #langchain #googlecloud
Dialogflow vs Vertex AI Conversation vs Agent builder... Making sense of the madness (in 10 minutes)
Просмотров 6 тыс.3 месяца назад
(..and 20 seconds I guess.) In this video I break down the differences between Dialogflow ES, Dialogflow CX, Vertex AI Conversation, Playbooks, Agent builder (formerly known as Gen App Builder), etc. (yeah, I know, it's a confusing landscape). Timestamps: 00:00 Intro 00:20 History of Dialogflow 01:02 Generative playbooks 01:51 Chat apps (fka Vertex AI Conversation) 03:28 Generative AI settings ...
Imagen 3 is very nice I think
Просмотров 4563 месяца назад
An in-depth look at Imagen 3, its image generation capabilities, benchmarks, and how to use it. note: yeah, I didn't notice the repeated "no" in the text with the Alien image. Tricky bastard. Colab notebook: colab.research.google.com/drive/1TSNDfzGCbH2ymaSmwhokvyFVtkno3jOv#scrollTo=-eAV7xVWZCpf Image FX: aitestkitchen.withgoogle.com/tools/image-fx Availability varies among countries and among w...
Google AI Studio vs Vertex AI vs Gemini... Making sense of the madness
Просмотров 17 тыс.4 месяца назад
UPDATE: Google AI Studio pricing is now lower than what I showed in the video and matches that of Vertex AI. This video is a breakdown of the Google AI landscape, specifically the different ways you can access Gemini models. I talk about Google AI Studio and Google for Developers, Vertex AI (and Vertex AI Studio), Gemini (Advanced vs Free), their differences, who they're for, pricing, and more....
Run Serverless LLMs with Ollama and Cloud Run (GPU Support)
Просмотров 1,5 тыс.4 месяца назад
A quick overview of the recently announced GPU support on Cloud Run and a walkthrough of deploying an Ollama container to server Gemma2-9B. Step by step tutorial (incl. Dockerfile and all required configs) exactly as shown in the video: cloud.google.com/run/docs/tutorials/gpu-gemma2-with-ollama Preview sign up form: services.google.com/fb/forms/cloudrungpusignup/
The rundown on the current state of Gemini models on Google Cloud
Просмотров 1465 месяцев назад
The rundown on the current state of Gemini models on Google Cloud
Agent Builder with BigQuery (SQL Agent on Google Cloud) - Tutorial
Просмотров 5 тыс.6 месяцев назад
Agent Builder with BigQuery (SQL Agent on Google Cloud) - Tutorial
Context Caching for Gemini on Vertex AI (Save up to 75% on input tokens)
Просмотров 4306 месяцев назад
Context Caching for Gemini on Vertex AI (Save up to 75% on input tokens)
Claude 3.5 Sonnet vs Gemini Advanced (Gemini 1.5 Pro) - Comparison
Просмотров 2,2 тыс.7 месяцев назад
Claude 3.5 Sonnet vs Gemini Advanced (Gemini 1.5 Pro) - Comparison
Run Langchain Apps on Google Cloud (Reasoning Engine API Walkthrough and Sample Code)
Просмотров 8327 месяцев назад
Run Langchain Apps on Google Cloud (Reasoning Engine API Walkthrough and Sample Code)
Google Cloud Agent Builder - Full Walkthrough (Tutorial)
Просмотров 24 тыс.8 месяцев назад
Google Cloud Agent Builder - Full Walkthrough (Tutorial)
Vertex AI Grounding with Google Search (Next 24 announcement)
Просмотров 4849 месяцев назад
Vertex AI Grounding with Google Search (Next 24 announcement)
Gemini has a memory function (not really, but sort of)
Просмотров 5369 месяцев назад
Gemini has a memory function (not really, but sort of)
A new (simpler) way to create Gen AI chatbots in Vertex AI: Generative Playbook Agents
Просмотров 3,8 тыс.10 месяцев назад
A new (simpler) way to create Gen AI chatbots in Vertex AI: Generative Playbook Agents
Gemini Advanced vs ChatGPT Plus Comparison
Просмотров 19 тыс.11 месяцев назад
Gemini Advanced vs ChatGPT Plus Comparison
Build your own Gemini Web app with Streamlit (Sample code included)
Просмотров 1,3 тыс.Год назад
Build your own Gemini Web app with Streamlit (Sample code included)
A first look at Gemini capabilities (Google AI Studio, Vertex AI Studio, and sample Notebook) - Demo
Просмотров 365Год назад
A first look at Gemini capabilities (Google AI Studio, Vertex AI Studio, and sample Notebook) - Demo
Bard (Gemini Pro) vs ChatGPT (GPT 3.5) - Quick comparison
Просмотров 4,5 тыс.Год назад
Bard (Gemini Pro) vs ChatGPT (GPT 3.5) - Quick comparison
Fine-tuning Codey foundation model (code-bison) on Google Cloud
Просмотров 782Год назад
Fine-tuning Codey foundation model (code-bison) on Google Cloud
Generative AI on Vertex AI: Model Garden and Generative AI Studio - Overview and DEMO
Просмотров 1,1 тыс.Год назад
Generative AI on Vertex AI: Model Garden and Generative AI Studio - Overview and DEMO
Building an Enterprise Search app using Vertex AI Search on Google Cloud (Demo)
Просмотров 10 тыс.Год назад
Building an Enterprise Search app using Vertex AI Search on Google Cloud (Demo)
Working with GKE on Google Cloud Platform - Demo Walkthrough
Просмотров 174Год назад
Working with GKE on Google Cloud Platform - Demo Walkthrough
Duet AI on Google Cloud - Quick Overview and How to Sign Up
Просмотров 683Год назад
Duet AI on Google Cloud - Quick Overview and How to Sign Up
A first look at Vertex AI Conversation (aka Gen App Builder Conversational AI) - Live demo
Просмотров 14 тыс.Год назад
A first look at Vertex AI Conversation (aka Gen App Builder Conversational AI) - Live demo