1 00:00:00,000 --> 00:00:04,000 So guys, we are going to continue the discussion with respect to Lange chain and hugging face integration. 2 00:00:04,000 --> 00:00:09,000 Um, in this video I'm going to show you an end to end project with the help of hugging face endpoint. 3 00:00:10,000 --> 00:00:17,000 Uh, if you remember, guys, uh, previously, uh, there was one amazing, um, end to end projects 4 00:00:17,000 --> 00:00:18,000 that we had actually created. 5 00:00:18,000 --> 00:00:18,000 Right? 6 00:00:18,000 --> 00:00:24,000 And if I talk about that specific project, it was about text summarization, where we were specifically 7 00:00:24,000 --> 00:00:28,000 using a YouTube URL or any website URL. 8 00:00:28,000 --> 00:00:32,000 It was able to summarize the entire text and give it to us, right? 9 00:00:32,000 --> 00:00:34,000 So I hope everybody remembers this. 10 00:00:34,000 --> 00:00:35,000 If you don't remember it, don't worry. 11 00:00:35,000 --> 00:00:42,000 I have copied the same content inside my ninth folder hugging face with lang chain in my app dot Pi. 12 00:00:42,000 --> 00:00:46,000 Okay, so let me just go ahead and run this quickly and then we will see that how we can. 13 00:00:46,000 --> 00:00:49,000 And over here you can see that I'm calling grok API key. 14 00:00:49,000 --> 00:00:53,000 And the model that is specifically used is uh gamma seven B. 15 00:00:53,000 --> 00:00:59,000 Right now instead of using gamma seven B, what I will do is that I will use my model from hugging face. 16 00:00:59,000 --> 00:01:00,000 Right. 17 00:01:00,000 --> 00:01:06,000 And that is also the same model that we used it over here, right, in which we discussed in our previous 18 00:01:06,000 --> 00:01:06,000 video. 19 00:01:06,000 --> 00:01:09,000 So let's first of all go this go and see this particular demo. 20 00:01:09,000 --> 00:01:12,000 So I am in my ninth hugging face with Lang chain folder. 21 00:01:12,000 --> 00:01:19,000 I will go ahead and right Streamlit run app dot pi okay so let's go ahead and see this. 22 00:01:19,000 --> 00:01:21,000 So here you will be able to see that. 23 00:01:21,000 --> 00:01:22,000 Fine. 24 00:01:22,000 --> 00:01:23,000 Uh this is my entire model. 25 00:01:23,000 --> 00:01:30,000 Or let me just copy this quickly and let me go back to my page over here. 26 00:01:30,000 --> 00:01:31,000 I will try to load it. 27 00:01:31,000 --> 00:01:31,000 Okay. 28 00:01:31,000 --> 00:01:34,000 So here is my entire page here. 29 00:01:34,000 --> 00:01:36,000 First of all, it is asking me for the grok API key. 30 00:01:36,000 --> 00:01:42,000 So I will go ahead and take my grok API key which is present in my env file. 31 00:01:42,000 --> 00:01:46,000 Sorry, grok API key, not a token, but grok API key is required. 32 00:01:46,000 --> 00:01:50,000 So I will go back over here, copy and paste it and press enter. 33 00:01:50,000 --> 00:01:50,000 Okay. 34 00:01:50,000 --> 00:01:55,000 Then I will take this particular URL okay AI versus ML versus generative AI. 35 00:01:55,000 --> 00:01:59,000 This is my video in my YouTube channel just to test it out. 36 00:02:00,000 --> 00:02:05,000 And I'll paste this link over here and I'll say, hey, summarize the content from the YouTube or website. 37 00:02:05,000 --> 00:02:11,000 And remember, right now we are using gamma two model from the um, from the API itself, from the grok 38 00:02:11,000 --> 00:02:12,000 API. 39 00:02:12,000 --> 00:02:13,000 And here you can see the entire output. 40 00:02:13,000 --> 00:02:19,000 Similarly, if you go ahead and see, uh, any other website page, right. 41 00:02:19,000 --> 00:02:24,000 Like let's say docs underscore dot dot dot lang chain.com. 42 00:02:24,000 --> 00:02:28,000 If I'm giving this particular URL in my summarize URL. 43 00:02:28,000 --> 00:02:33,000 And if I say to summarize the content from the YouTube or website, you'll be able to see that I'm able 44 00:02:33,000 --> 00:02:35,000 to get the entire summary. 45 00:02:35,000 --> 00:02:39,000 Okay, so this is perfectly fine till here. 46 00:02:39,000 --> 00:02:40,000 I think you'll be able to get it right. 47 00:02:40,000 --> 00:02:41,000 So all the information is there. 48 00:02:41,000 --> 00:02:46,000 Now, instead of using this grok API and Google Gamma two model, I will try to use Huggingface endpoint 49 00:02:46,000 --> 00:02:46,000 again. 50 00:02:46,000 --> 00:02:48,000 It is completely for free. 51 00:02:48,000 --> 00:02:50,000 Uh, with respect to some of the models that we have. 52 00:02:50,000 --> 00:02:50,000 Okay. 53 00:02:50,000 --> 00:02:53,000 So I will go over here, go back to my API. 54 00:02:53,000 --> 00:02:57,000 I will just close this because I don't want it right now okay. 55 00:02:57,000 --> 00:02:59,000 And let's see what all changes I will do over here. 56 00:02:59,000 --> 00:03:03,000 First of all you you'll be seeing that this all will be same, right? 57 00:03:03,000 --> 00:03:06,000 First of all, I will go ahead and change this API. 58 00:03:06,000 --> 00:03:06,000 Right. 59 00:03:06,000 --> 00:03:12,000 Instead of writing grok API key, I will say, hey, uh, hugging face API token. 60 00:03:12,000 --> 00:03:15,000 Okay, I'll, I'll say provide my hugging face API token. 61 00:03:15,000 --> 00:03:16,000 Okay. 62 00:03:16,000 --> 00:03:18,000 And I will write HDF over here. 63 00:03:18,000 --> 00:03:19,000 Okay. 64 00:03:19,000 --> 00:03:25,000 So first thing is this then uh, you'll be able to see that I will also make sure that I will comment 65 00:03:25,000 --> 00:03:29,000 out this particular code, because LM model I need to call with respect to that. 66 00:03:29,000 --> 00:03:34,000 Now, if I really need to call the LM model from hugging face endpoint, I need to import this library 67 00:03:34,000 --> 00:03:38,000 from long chain underscore hugging face so quickly I will go over here. 68 00:03:38,000 --> 00:03:40,000 I will import it over here okay. 69 00:03:40,000 --> 00:03:45,000 Once I import it uh, now it's time that I should be calling my hugging face API. 70 00:03:45,000 --> 00:03:47,000 And let me just do one thing. 71 00:03:47,000 --> 00:03:52,000 Let me also make sure to put this particular condition over here for my API, hf underscore API. 72 00:03:52,000 --> 00:03:59,000 So I'll say if HF underscore API dot strip uh and here I'm using lm so lm now I have to create. 73 00:03:59,000 --> 00:04:00,000 Right. 74 00:04:00,000 --> 00:04:06,000 So I will go over here I have commented out this particular LM and instead of writing like this I will 75 00:04:06,000 --> 00:04:07,000 go to my experiment page. 76 00:04:07,000 --> 00:04:09,000 And you know that how I have called this. 77 00:04:09,000 --> 00:04:10,000 Right. 78 00:04:10,000 --> 00:04:14,000 So I will just copy this entire thing like Mistral seven B and all. 79 00:04:14,000 --> 00:04:15,000 I will paste it over here. 80 00:04:16,000 --> 00:04:17,000 That's it. 81 00:04:17,000 --> 00:04:18,000 That is the change. 82 00:04:19,000 --> 00:04:21,000 So simple right now. 83 00:04:21,000 --> 00:04:24,000 Your task will be that whatever end to end projects we have created, right? 84 00:04:24,000 --> 00:04:30,000 Just try to see whether we are having something or not related to this right now. 85 00:04:30,000 --> 00:04:35,000 Inside this token, you will be able to see that I'm calling OS dot get env right h f underscore token 86 00:04:35,000 --> 00:04:36,000 right. 87 00:04:36,000 --> 00:04:41,000 So this particular token needs to be replaced with whatever token I am going to get. 88 00:04:41,000 --> 00:04:44,000 So this is nothing, but it will be a hf API key, right? 89 00:04:44,000 --> 00:04:46,000 Whatever token I will be getting from my website. 90 00:04:46,000 --> 00:04:47,000 That's it. 91 00:04:47,000 --> 00:04:48,000 Right? 92 00:04:49,000 --> 00:04:50,000 This is so simple. 93 00:04:50,000 --> 00:04:51,000 That's it. 94 00:04:51,000 --> 00:04:51,000 Right. 95 00:04:51,000 --> 00:04:57,000 And here you have actually done that change this entire model where you have calling this particular 96 00:04:57,000 --> 00:04:59,000 LLM model from the API with the help of API itself. 97 00:04:59,000 --> 00:05:02,000 So now let's go ahead and execute this. 98 00:05:02,000 --> 00:05:06,000 I'll say Streamlit run App.py. 99 00:05:06,000 --> 00:05:09,000 And now, finally, you can see that I've executed it. 100 00:05:10,000 --> 00:05:11,000 Now let's do the same thing. 101 00:05:11,000 --> 00:05:12,000 I will take this. 102 00:05:12,000 --> 00:05:12,000 Now. 103 00:05:12,000 --> 00:05:14,000 My LM model is completely changed. 104 00:05:14,000 --> 00:05:17,000 Before that, I'll just go ahead and put my hugging face API token. 105 00:05:17,000 --> 00:05:21,000 It is present over here in my env file. 106 00:05:21,000 --> 00:05:23,000 And here I will go back. 107 00:05:23,000 --> 00:05:24,000 I will paste it over here. 108 00:05:25,000 --> 00:05:29,000 Okay, now let's try the same YouTube URL. 109 00:05:29,000 --> 00:05:33,000 The content will be different when compared to Google Gamma two. 110 00:05:33,000 --> 00:05:35,000 Because Google Gamma two is a much better model. 111 00:05:35,000 --> 00:05:37,000 It is a huge 9 billion parameters, right? 112 00:05:37,000 --> 00:05:40,000 And this is just a smaller model that we have. 113 00:05:40,000 --> 00:05:41,000 But it is also a very good model. 114 00:05:41,000 --> 00:05:45,000 So here you can see in this video Kushner discusses the differences. 115 00:05:45,000 --> 00:05:46,000 It says Kushner okay. 116 00:05:46,000 --> 00:05:50,000 Whenever I say crush it assumes crush okay. 117 00:05:50,000 --> 00:05:53,000 And all the information is there with respect to the summary, the same thing. 118 00:05:53,000 --> 00:05:56,000 What I will do, I will also do it for this particular website. 119 00:05:56,000 --> 00:06:03,000 And here easily you could see that okay, so beautifully we are able to do all these things right now. 120 00:06:03,000 --> 00:06:05,000 I'll go ahead and summarize the content. 121 00:06:07,000 --> 00:06:08,000 Any page. 122 00:06:08,000 --> 00:06:11,000 You should be able to summarize it in a much more easy way. 123 00:06:11,000 --> 00:06:12,000 Right. 124 00:06:12,000 --> 00:06:17,000 So again this model somewhat time it is taking. 125 00:06:17,000 --> 00:06:19,000 But now you can see this entire information. 126 00:06:19,000 --> 00:06:20,000 Right. 127 00:06:20,000 --> 00:06:20,000 right? 128 00:06:20,000 --> 00:06:23,000 So yeah, this was it from my side. 129 00:06:23,000 --> 00:06:25,000 This was with respect to Lang Chen and Huggingface. 130 00:06:25,000 --> 00:06:27,000 I hope you like this particular video. 131 00:06:27,000 --> 00:06:32,000 And now one of the tasks that I really want to give, go ahead and explore other projects and try to 132 00:06:32,000 --> 00:06:36,000 integrate with Huggingface endpoint with using any models that you really want. 133 00:06:36,000 --> 00:06:37,000 Okay. 134 00:06:37,000 --> 00:06:41,000 See what kind of issues you face and then you can probably proceed ahead with that. 135 00:06:41,000 --> 00:06:42,000 So yes, this was it from my side. 136 00:06:42,000 --> 00:06:44,000 I hope you liked this particular video. 137 00:06:44,000 --> 00:06:45,000 I will see you all in the next video. 138 00:06:45,000 --> 00:06:45,000 Thank you. 139 00:06:45,000 --> 00:06:46,000 Take care. 140 00:06:46,000 --> 00:06:46,000 Bye bye.