WEBVTT 00:00.280 --> 00:01.560 Hey there, Ethan here. 00:01.560 --> 00:06.520 And in this video, we'll be setting up our project of our reflection agent. 00:06.840 --> 00:13.480 So we'll create the project directory, and we'll use poetry to create our virtual environment and to 00:13.520 --> 00:15.040 install our dependencies. 00:15.280 --> 00:21.440 We'll then open PyCharm and configure it to point to the virtual environment we are using. 00:22.200 --> 00:30.080 And we'll create a dot env file that will hold all of our open AI API keys and API keys, because we're 00:30.080 --> 00:32.280 going to be using Lindsmith for tracing. 00:35.280 --> 00:36.040 Alrighty. 00:36.040 --> 00:43.360 So let's go to desktop and I want to create a new directory and I'm going to call it Reflection agent. 00:44.920 --> 00:46.680 I'm going to CD into it. 00:47.120 --> 00:50.600 And now I want to initialize my portal environment. 00:50.600 --> 00:52.520 So I'm going to write poetry init. 00:52.960 --> 00:56.040 And I'm going to click enter for everything. 01:02.600 --> 01:05.680 And I'll have shown my poetry virtual environment ready. 01:06.120 --> 01:06.560 Cool. 01:06.560 --> 01:07.560 So we have it ready. 01:07.600 --> 01:09.880 We can take a look at our HTML file. 01:10.600 --> 01:12.400 And it's created. 01:12.400 --> 01:15.920 And now let's install the packages we need for this project. 01:16.320 --> 01:22.360 So I'll start by writing Poetry Add and I'm going to add Python dotenv to load environment variables 01:22.360 --> 01:23.720 from our dot env file. 01:24.280 --> 01:28.080 And I'm going to install black and eishort for formatting. 01:28.600 --> 01:31.560 And I now want to install link chain. 01:31.680 --> 01:37.760 I'll also install link chain OpenAI because we're going to be using the GPT 3.5 model. 01:37.960 --> 01:40.320 And of course let's install link graph. 01:42.000 --> 01:44.880 So this will take us a couple of seconds. 01:45.200 --> 01:48.560 Now the virtual environment is installing all the dependencies. 01:53.240 --> 01:54.880 And we have everything ready. 01:55.360 --> 01:58.400 So now let's go and open PyCharm. 01:58.880 --> 02:02.880 And we want now to open a new project. 02:03.320 --> 02:07.160 And I'm going to find the new directory I created. 02:07.520 --> 02:11.640 And PyCharm is going to detect my poetry virtual environment. 02:12.040 --> 02:18.400 And we can see we have our pyproject.toml file and our poetry dot log file, which has all the exact 02:18.400 --> 02:20.920 versions of the packages that we're using. 02:21.400 --> 02:21.960 Alrighty. 02:22.040 --> 02:24.480 let's go create now a dot env file. 02:26.480 --> 02:30.720 And now I'm going to paste some environment variables we'll need for this project. 02:31.080 --> 02:33.400 Now I assume you know how to get them. 02:33.560 --> 02:35.400 We're going to be using OpenAI. 02:35.440 --> 02:37.440 So we need the OpenAI API key. 02:37.720 --> 02:40.920 And we're going to be using Lamb-smith for tracing. 02:41.040 --> 02:43.480 So we want to get the API key. 02:43.760 --> 02:46.600 We'll set the length chain tracing v2 equals true. 02:46.920 --> 02:51.520 And let's set our length chain project the display name to be the reflection agent. 02:56.160 --> 02:56.800 Alrighty. 02:56.800 --> 02:59.320 Let's go and create a new Python file. 02:59.320 --> 03:01.520 And this is going to be our main file. 03:02.000 --> 03:06.600 And let's just write the boilerplate code of if name equals main. 03:08.680 --> 03:11.640 Then we want to print hello link graph. 03:12.800 --> 03:15.360 Let's run it as a sanity check. 03:17.840 --> 03:20.040 And we can see that this is working. 03:20.360 --> 03:20.760 Yay! 03:21.720 --> 03:27.360 Alrighty let's go and import from env the load dot n function. 03:28.440 --> 03:30.120 And we want to call it. 03:30.120 --> 03:34.330 So this will take all the environment variables from the dot dotenv file and load them. 03:34.370 --> 03:36.250 Let's just run this in debug. 03:36.890 --> 03:42.010 And let's see that we actually get values of those environment variables. 03:43.490 --> 03:46.970 So I'm going to simply here evaluate an expression. 03:46.970 --> 03:49.730 Let's import OS enter. 03:49.930 --> 03:51.450 So now we've imported it. 03:51.730 --> 03:55.290 And let's now check out the OpenAI API key. 03:55.530 --> 03:58.930 So I'm going to root OS dot environ. 03:59.250 --> 04:03.610 And I'm going to reference the key of OpenAI API key. 04:04.290 --> 04:04.730 Cool. 04:04.770 --> 04:07.130 We see the value has been successfully loaded. 04:07.610 --> 04:13.210 And before we go and implement our reflection agent, let's go and check out the poetry dot log file. 04:13.530 --> 04:16.690 And let's check out the versions of the packages we're using. 04:16.930 --> 04:18.610 So I'm going to search for link chain. 04:18.650 --> 04:21.970 I have 0.1 16 which is the current latest. 04:21.970 --> 04:27.650 And if I look for line graph we have 0.38 which is currently the latest. 04:28.010 --> 04:31.890 And we can see the declaration in the pyproject.toml file. 04:32.250 --> 04:39.050 And we're done setting up our boilerplate project with all the dependencies we need all the environment 04:39.050 --> 04:40.330 variables loaded. 04:40.330 --> 04:43.290 And now it's time to write some line graph code.