WEBVTT 1 00:00:00.150 --> 00:00:02.220 Welcome to this course. 2 00:00:02.220 --> 00:00:04.470 My name is Maximilian Schwarzmuller. 3 00:00:04.470 --> 00:00:06.540 I will be your instructor in this course 4 00:00:06.540 --> 00:00:10.320 and in this course we'll dive deep into running 5 00:00:10.320 --> 00:00:14.040 openAI models locally on your system, your laptop, 6 00:00:14.040 --> 00:00:17.520 your desktop PC, on your server if you want to, 7 00:00:17.520 --> 00:00:19.560 though the focus of this course will not be 8 00:00:19.560 --> 00:00:22.020 server administration, but of course you could apply 9 00:00:22.020 --> 00:00:23.460 the knowledge you gain in this course 10 00:00:23.460 --> 00:00:25.650 to also run open models there. 11 00:00:25.650 --> 00:00:28.980 Because in this course, we will explore in depth 12 00:00:28.980 --> 00:00:32.350 how we can use tools like LM Studio or Ollama 13 00:00:33.418 --> 00:00:36.420 to run openAI models, large language models 14 00:00:36.420 --> 00:00:39.870 like the Gemma models by Google on any hardware, 15 00:00:39.870 --> 00:00:42.510 including consumer-level hardware. 16 00:00:42.510 --> 00:00:45.060 You will learn about hardware requirements 17 00:00:45.060 --> 00:00:48.300 and how you can run quite powerful models, 18 00:00:48.300 --> 00:00:51.420 even on low-end devices, how that's possible. 19 00:00:51.420 --> 00:00:54.780 We'll take a close look at that and of course you will learn 20 00:00:54.780 --> 00:00:57.480 how to leverage LM Studio and Ollama, 21 00:00:57.480 --> 00:01:00.540 the tools we'll be using to run these AI models. 22 00:01:00.540 --> 00:01:03.120 You will learn about advanced configuration options. 23 00:01:03.120 --> 00:01:06.960 You'll see plenty of examples, and of course we'll also dive 24 00:01:06.960 --> 00:01:11.960 into example use cases for using open models on your system, 25 00:01:12.210 --> 00:01:14.220 because the truth is that 26 00:01:14.220 --> 00:01:17.550 these AI models are extremely powerful. 27 00:01:17.550 --> 00:01:21.390 We're no longer in 2023 where you essentially 28 00:01:21.390 --> 00:01:23.370 had to go for paid models. 29 00:01:23.370 --> 00:01:27.300 Instead now these open models, which you can run for free, 30 00:01:27.300 --> 00:01:29.850 are a very viable alternative 31 00:01:29.850 --> 00:01:32.220 to these paid proprietary models, 32 00:01:32.220 --> 00:01:35.730 and they have some real advantages, like for example, 33 00:01:35.730 --> 00:01:40.170 100% privacy because your prompts, the generated output 34 00:01:40.170 --> 00:01:42.420 and any data you use in your prompts 35 00:01:42.420 --> 00:01:46.050 never leaves your machine or your network. 36 00:01:46.050 --> 00:01:48.690 But we'll take a close look at all that throughout 37 00:01:48.690 --> 00:01:51.690 this course, and therefore, let's not waste any time. 38 00:01:51.690 --> 00:01:53.250 Let's instead get started 39 00:01:53.250 --> 00:01:55.830 and let's start with the most important question 40 00:01:55.830 --> 00:01:58.440 to make sure that we're all on the same page. 41 00:01:58.440 --> 00:02:01.260 What exactly does open model 42 00:02:01.260 --> 00:02:04.050 or open Large Language Model mean? 43 00:02:04.050 --> 00:02:06.723 What are these open models?