**Playing World of Warcraft Classic on the GeForce RTX 490**

I'm trying to play World of Warcraft Classic on my 4090, and I want to know if that's going to work out good for me. The GeForce RTX 490 is a powerful GPU, and it should provide a great gaming experience for World of Warcraft classic with high frame rates and detailed graphics.

**Tanking Ragnaros in Molten Core**

I'm trying to tank Ragnaros in Molten Core, and I'd like some advice on what kind of resistance to prioritize. I'd recommend prioritizing fire resistance, as Ragnaros deals significant fire damage. If you're looking for a good weapon drop from the fight, the Sulfuras, Unopened, is likely to be it - a high-damage two-handed mace.

**Mr. Bigglesworth and the Digital Human**

I've been doing some research on digital human blueprints, and I came across Mr. Bigglesworth, who seems to be a boss in Molten Core. I'm not familiar with this character, but it sounds like he's an interesting addition to the game. The idea of having a digital human that can interact with players is also pretty cool.

**Agent Blueprints**

I've been exploring agent blueprints and have some ideas for how they could be used in the future. One potential direction I'd like to explore is using vision language models, such as VLMs, to enhance the digital human's abilities. This could allow the digital human to use the camera to show players something or upload files. The idea of having a digital human that can pick up on emotions from facial expressions also seems like it could be pretty cool.

**Digital Human Repository**

For those interested in learning more about the digital human project, I've created a repository on GitHub that includes all of the blueprints and code for the project. This should provide a good starting point for anyone who wants to get involved with the project.

**Next Steps**

I'm excited to see where this project is going to go next. If you enjoy learning about digital humans and agent blueprints, be sure to follow along on this journey. There's still a lot of work to be done, but I think it has the potential to be really cool.

"WEBVTTKind: captionsLanguage: enhello James can you explain to the audience what Nvidia Nim is NVIDIA Nim is a suite of inference microservices that simplifies the deployment of AI models at scale providing optimized containers for popular AI models and enabling Enterprises to maximize their infrastructure Investments okay that's pretty cool can you also explain what Nvidia Nim agent blueprints is NVIDIA NM agent blueprints are preconfigured reusable templates that streamline the creation of nim agents allowing developers to quickly deploy and manage AI models in production environments okay that sounds pretty cool uh I'm going to try that out so thank you for your help and have a great day you're welcome I'm glad I could help and I hope you have a great experience with Nvidia Nim okay so what we just heard from James here today we are taking a look at what are Nvidia Nim agent blueprints so you can see here these agent blueprints are reference workflows for canical generative a use cases so enter can build an operationalize custom AI applications creating data driven AI flyways using n anent blueprints along with the Nvidia Nim microservices that we have uh played around with on this channel before so I thought we to going to look at the three different blueprints we have available now so we had a look at the digital human we have this rag setup here with PDF data extraction and we have this generary virtual screening for drug Discovery so I thought we can dive into these three uh different blueprints today take a quick look at them try them out and see how they work so yeah let's just get going so if you want to try this out now you can just follow the link in the description uh we can go to build nvidia.com right and we will find this explore Nim agent blueprints uh view all hent blueprints right and here we can see we have three different ones we can try out now so if we take a look at this multimodal PDF extraction we kind of have uh some files we can choose from here so these are pre-up loaded from Nvidia that's we can try out right and you can see we have the Nim here so we can look at the setup so we are using like Google deploy uh Nvidia rerank mols this is like a a rag ranking system we have the embeddings model so you can kind of dive into what we are using on the back end here right so let me show you how this works now so let's say we uh go to public earnings 370 files and you can see we can click on extract data set this is loading up our data set it's ready to test right uh so we can ask any question here so let me just come up with some question okay so let's do how much money did apple make on iPhone sales in 2023 so you can see this should be pretty quick okay so you can see apple made two is that 200 billion I think so or two uh on iPhone sales in 2023 and what's nice here now we can see we have this uh resources here we can look look up right so I think that's pretty neat uh I'm going to zoom in on this but you can see here the revenue right is 200 million so that's up from 2022 so that was pretty quick let's do another quick question here so let's do in the last three years what years sold most Max descending order so let's see if we can get our response on that uh 2022 had the highest followed by 21 then 23 so we can check that okay so I guess you barely can see it so 22 was 40 billion 21 was 35 and 23 was 29 billion so yeah uh pretty good uh now let's move on and just try another blueprint shall we okay so let's go into this generative virtual screening pipeline so this is way over my head but let's just try it now so you can see we have Alpha fold 2 that's pretty cool that is the big Google folded proteins model uh I'm not going to go into any specific here but let's just try to change up some inputs here and Target Protein that's too hard for me so let's just try to click here generate molecules uh because I think this has a pretty cool interface right so let's just place let this cook for a while and I'll take you back when we can see this okay so now you can see we kind of have our molecule here we can zoom in about on it right and we can kind of look in all directions so this must be pretty cool if you are working on this right look at this I don't know even what this is it's some kind of protein I guess uh yeah I really think the interface here is pretty cool and we can kind of look at these different molecules right we can zoom in on them and there are different stuff here we can look at but again way over my head amino acid sequence yeah uh this is nothing I can go into dep but we have the option here maybe professionals can log into this agent blueprint get up right away and start using this so yeah uh let's move on I want to take a bit more close look at the human digital human that is kind of my favorite so far so I guess you kind of saw me use this on the intro but you can see we have some different options here so yeah just turn off this dark reader so we can deselect or select nvidia's producting market knowledge base so I use that in intro right but if we deselect this and then I think we can just talk to James uh without using that right so let's just click to speak and just do like a short conversation and then I'm going to show you kind of the GitHub we have access to with this Nvidia and name ENT Blueprints and kind of move on to that so yeah let's talk to James hello James so I'm trying to play World of Warcraft Classic on my 4090 is that going to work out good for me the GeForce RTX 490 is a powerful GPU and it should provide a great gaming experience for World of Warcraft classic with high frame rates and detail Graphics okay that sounds good so I'm trying also to tank r raros in molon core what resistance should I prioritize for tanking ragnaros in molten core I'd recommend prioritizing fire resistance as he deals significant fire damage okay I'll go ahead and do that what kind of weapon is the best drop from Mr ragnaros the best weapon drop from Mr Bigglesworth is likely the sulfuras the unopened high damage two-handed mace right okay Mr bigles word who is that okay I never done that Mr Bigglesworth is a boss in Molton core a raid instance in World of Warcraft classic okay I haven't done him but thank you anyway have a great day and we speak soon right you're welcome have a great day and I'll look forward to speaking with you soon bye bye okay so Mr bigo s yeah that's a cat in AAS I totally forgot about that but that was pretty funny it got the fire it's correct right so that was pretty cool uh I think that's this is pretty cool I like this and right I have a I'm going to share like a few ideas I have around this uh agent blueprint that I might try out in the future right uh but let's dive over to the GitHub repo I'm going to leave a link in the description for this because here you also have access to all the blueprints here so you can see we have this repo so we can go into the digital human repo here uh we have the architecture if you want to take a look at that and here we kind of have a get started prerequisite all you have to do we can also go back here let's say the digital human we can click on this Blueprint Card and here we can kind of see the system requirements so this is a compute heavy pipeline 2 times a100 yeah so you kind of need like a some compute to actually run this digital human right but we have a good uh installation here so we're going to run like a Docker GPA we're going to run this uh GPU driver version 4 3 we have the install the container Cool kit so you can follow along here and if you go back here you can see you can access all of these different GitHub repos if you wanted to set this up for yourself yeah so you can see here again this is the multimodal one so I guess this has less requirements right so a lot to play around with here if you want to check it out so as you probably understand by now my favorite blueprint is the digital human so I had some different ideas around here that maybe we can use in the future right so here are kind of the names we are using for is now we have the audio to face uh that is the video we see from James with the lip syncing right and the facial expressions and we run kind of llama 38b uh in the back end and we also have this that kind of transfers our audio to text for uh llama 8B right but let's say we wanted to expand on this digital hum so what I thought about was it would be pretty cool if we can add like this vlm so that's Vision language models like let's say we could swap out llama B and we can also introduce like a new name here that can take Vis in right so we can use the camera and let's say I could use the camera here to kind of show uh our digital humans something right or upload some kind of files uh I thought that could be pretty cool and you can kind of see my facial expression so I'm hoping for something like that in the future and that kind of will bring the realism to another level and I think like in games uh let's say the human or like the digital human can kind of see your facial expression too and kind of pick up on some emotions in your face I I think that could be pretty cool so yeah if you want to try this out just follow the link in the description it will take you over to here and you can kind of uh I would really recommend try talking with this digital human it's pretty cool if you ask me and also check out the other ancient blueprint here head over to GitHub if you want to explore even more into the code and stuff so yeah pretty cool if you ask me looking forward to see uh where the next step of this Nvidia n micro Services is going to be so if you enjoy this give this video a like thank you for tuning in have a great day and we speak soonhello James can you explain to the audience what Nvidia Nim is NVIDIA Nim is a suite of inference microservices that simplifies the deployment of AI models at scale providing optimized containers for popular AI models and enabling Enterprises to maximize their infrastructure Investments okay that's pretty cool can you also explain what Nvidia Nim agent blueprints is NVIDIA NM agent blueprints are preconfigured reusable templates that streamline the creation of nim agents allowing developers to quickly deploy and manage AI models in production environments okay that sounds pretty cool uh I'm going to try that out so thank you for your help and have a great day you're welcome I'm glad I could help and I hope you have a great experience with Nvidia Nim okay so what we just heard from James here today we are taking a look at what are Nvidia Nim agent blueprints so you can see here these agent blueprints are reference workflows for canical generative a use cases so enter can build an operationalize custom AI applications creating data driven AI flyways using n anent blueprints along with the Nvidia Nim microservices that we have uh played around with on this channel before so I thought we to going to look at the three different blueprints we have available now so we had a look at the digital human we have this rag setup here with PDF data extraction and we have this generary virtual screening for drug Discovery so I thought we can dive into these three uh different blueprints today take a quick look at them try them out and see how they work so yeah let's just get going so if you want to try this out now you can just follow the link in the description uh we can go to build nvidia.com right and we will find this explore Nim agent blueprints uh view all hent blueprints right and here we can see we have three different ones we can try out now so if we take a look at this multimodal PDF extraction we kind of have uh some files we can choose from here so these are pre-up loaded from Nvidia that's we can try out right and you can see we have the Nim here so we can look at the setup so we are using like Google deploy uh Nvidia rerank mols this is like a a rag ranking system we have the embeddings model so you can kind of dive into what we are using on the back end here right so let me show you how this works now so let's say we uh go to public earnings 370 files and you can see we can click on extract data set this is loading up our data set it's ready to test right uh so we can ask any question here so let me just come up with some question okay so let's do how much money did apple make on iPhone sales in 2023 so you can see this should be pretty quick okay so you can see apple made two is that 200 billion I think so or two uh on iPhone sales in 2023 and what's nice here now we can see we have this uh resources here we can look look up right so I think that's pretty neat uh I'm going to zoom in on this but you can see here the revenue right is 200 million so that's up from 2022 so that was pretty quick let's do another quick question here so let's do in the last three years what years sold most Max descending order so let's see if we can get our response on that uh 2022 had the highest followed by 21 then 23 so we can check that okay so I guess you barely can see it so 22 was 40 billion 21 was 35 and 23 was 29 billion so yeah uh pretty good uh now let's move on and just try another blueprint shall we okay so let's go into this generative virtual screening pipeline so this is way over my head but let's just try it now so you can see we have Alpha fold 2 that's pretty cool that is the big Google folded proteins model uh I'm not going to go into any specific here but let's just try to change up some inputs here and Target Protein that's too hard for me so let's just try to click here generate molecules uh because I think this has a pretty cool interface right so let's just place let this cook for a while and I'll take you back when we can see this okay so now you can see we kind of have our molecule here we can zoom in about on it right and we can kind of look in all directions so this must be pretty cool if you are working on this right look at this I don't know even what this is it's some kind of protein I guess uh yeah I really think the interface here is pretty cool and we can kind of look at these different molecules right we can zoom in on them and there are different stuff here we can look at but again way over my head amino acid sequence yeah uh this is nothing I can go into dep but we have the option here maybe professionals can log into this agent blueprint get up right away and start using this so yeah uh let's move on I want to take a bit more close look at the human digital human that is kind of my favorite so far so I guess you kind of saw me use this on the intro but you can see we have some different options here so yeah just turn off this dark reader so we can deselect or select nvidia's producting market knowledge base so I use that in intro right but if we deselect this and then I think we can just talk to James uh without using that right so let's just click to speak and just do like a short conversation and then I'm going to show you kind of the GitHub we have access to with this Nvidia and name ENT Blueprints and kind of move on to that so yeah let's talk to James hello James so I'm trying to play World of Warcraft Classic on my 4090 is that going to work out good for me the GeForce RTX 490 is a powerful GPU and it should provide a great gaming experience for World of Warcraft classic with high frame rates and detail Graphics okay that sounds good so I'm trying also to tank r raros in molon core what resistance should I prioritize for tanking ragnaros in molten core I'd recommend prioritizing fire resistance as he deals significant fire damage okay I'll go ahead and do that what kind of weapon is the best drop from Mr ragnaros the best weapon drop from Mr Bigglesworth is likely the sulfuras the unopened high damage two-handed mace right okay Mr bigles word who is that okay I never done that Mr Bigglesworth is a boss in Molton core a raid instance in World of Warcraft classic okay I haven't done him but thank you anyway have a great day and we speak soon right you're welcome have a great day and I'll look forward to speaking with you soon bye bye okay so Mr bigo s yeah that's a cat in AAS I totally forgot about that but that was pretty funny it got the fire it's correct right so that was pretty cool uh I think that's this is pretty cool I like this and right I have a I'm going to share like a few ideas I have around this uh agent blueprint that I might try out in the future right uh but let's dive over to the GitHub repo I'm going to leave a link in the description for this because here you also have access to all the blueprints here so you can see we have this repo so we can go into the digital human repo here uh we have the architecture if you want to take a look at that and here we kind of have a get started prerequisite all you have to do we can also go back here let's say the digital human we can click on this Blueprint Card and here we can kind of see the system requirements so this is a compute heavy pipeline 2 times a100 yeah so you kind of need like a some compute to actually run this digital human right but we have a good uh installation here so we're going to run like a Docker GPA we're going to run this uh GPU driver version 4 3 we have the install the container Cool kit so you can follow along here and if you go back here you can see you can access all of these different GitHub repos if you wanted to set this up for yourself yeah so you can see here again this is the multimodal one so I guess this has less requirements right so a lot to play around with here if you want to check it out so as you probably understand by now my favorite blueprint is the digital human so I had some different ideas around here that maybe we can use in the future right so here are kind of the names we are using for is now we have the audio to face uh that is the video we see from James with the lip syncing right and the facial expressions and we run kind of llama 38b uh in the back end and we also have this that kind of transfers our audio to text for uh llama 8B right but let's say we wanted to expand on this digital hum so what I thought about was it would be pretty cool if we can add like this vlm so that's Vision language models like let's say we could swap out llama B and we can also introduce like a new name here that can take Vis in right so we can use the camera and let's say I could use the camera here to kind of show uh our digital humans something right or upload some kind of files uh I thought that could be pretty cool and you can kind of see my facial expression so I'm hoping for something like that in the future and that kind of will bring the realism to another level and I think like in games uh let's say the human or like the digital human can kind of see your facial expression too and kind of pick up on some emotions in your face I I think that could be pretty cool so yeah if you want to try this out just follow the link in the description it will take you over to here and you can kind of uh I would really recommend try talking with this digital human it's pretty cool if you ask me and also check out the other ancient blueprint here head over to GitHub if you want to explore even more into the code and stuff so yeah pretty cool if you ask me looking forward to see uh where the next step of this Nvidia n micro Services is going to be so if you enjoy this give this video a like thank you for tuning in have a great day and we speak soon\n"