Видео

You are the user. The AI is the assistant.

@@videotronicmaker lol... I know that - if you have LM Studio, make sure you upgrade to the most recent version and go to the chat, you will see that there are now "User" and "Assistant" mode you can input with both. Previously "System" role: Use this field to provide background instructions to the model, such as a set of rules, constraints, or general requirements.

I took a peak. I will look at it more later. Looks good. Thx.

openai==0.28.0 Here is the GitHub repo: github.com/VideotronicMaker/LM_Studio_Local_Server

Can anyone tell me

I am not sure which labels you are referring to exactly. There are several. Here is the link to the Uno Pinout Diagram. Most of the labels are the same: docs.arduino.cc/retired/boards/arduino-uno-rev3-with-long-pins/#osh-schematics If you want the labels for the others sections of the shield you can look at my other two videos that go deeper into the shields to get a better sense of how to use the shield and what each section of the shield may be. ruclips.net/video/gOo70ULv-yI/видео.html ruclips.net/video/czGsopX2CKI/видео.html I don't believe that Arduino made a diagram for this shield. That is why I originally made the video. I hope this helps a little. I was hoping someone else would jump in. Let me know exactly which labels you are referring to.

Yes. It sounds like you haven't installed the Openai library. 2 solutions: 1- go to my GitHub repo and run the requirements.txt file 2- Alternatively you can just use an AI assistant. Whenever you get errors just copy them and paste them into your AI assistant and it will tell you what to do.

I recently broke my main PC and I have been preoccupied with figuring out how to use non quantized models on my old MacBook Pro. I have still been stuck on Gemma 2b and Stable LM Zephyr 1.6b. I tried to inference the base Gemma 9b with no success. I hope to inference it very soon. I will say that because I was very impressed with 2b, I am sure that 9b and 27b will be impressive. I am gpu poor these days so...In time. Also, Google has really started to impress me lately with Gemini. All good things! I will be sure to edit this in a few days after I get a chance to try them. I guess I will try it out on Vertex AI in the meantime. Thanks for the question. I now have a next project to get into!

The model used in this video is the Phi 2 model by Microsoft. This model only understands English. Here is the huggingface page with more information. huggingface.co/microsoft/phi-2 You would need a different model for Spanish. I am not sure which models this size understand Spanish. However, I have seen models where other people have added a second version of the base model and made it in a different language so I would say check on hugging face and search for something like, "Phi-3-sp".

Thank you for commenting. I want to both encourage you and make something clear. I believe that all you need to implement or create this from scratch and on your own, is to utilize something like Chat GPT 4 or Google Gemini Advanced. That is exactly how I did it. I am not a coder, mathematician, computer science major or anything like that. My profession is television production and I also do a little web design. I was only able to do this by going back and forth with GPT 4. I gave it some code I found in another video and asked it to give me variations of that. Then, step by step I asked it to add an additional feature. It gave me the code. I pasted the code and usually there was an error. I then copied and pasted that error back into GPT 4 until it gave me code that worked. I knew nothing about python environments, anaconda, libraries or dependencies. When those items came up I asked it to explain. It did and I learned and moved to the next step. I must say that as of the release of GPT 4 o, I had a difficult time with GPT 4 o's performance so I stopped using it. I started using Gemini Advanced and to my surprise it has been great. I do not know how to code in Python. One thing I wanted people to get from this is that AI is amazing because it enables you to complete projects like this one. Use it. Contrary to what some say, AI is very useful beyond writing emails, planning trips and shopping. Also, I am using Gemini Advanced to tutor me in Python. I ask it to start simple and take me step by step and be my instructor. It works. That is what this channel is about. I show people what I am learning as I learn it. It is amazing to me so that is why I tell others. In this video, even though the AI is writing the code, I am learning little by little in the process. It may not be the most common approach but it is working for me. I hope this helps.

@@videotronicmaker Thank you for the excellent and lengthy answer that I will read with time, it's a true class what you did. I will follow your advice. I'm also a video producer and a musician. I know a little bit of computers and coding, but too little. Thanks. 🙏👍💥❤️

LM Studio is for running an offline llm. There is no connection to the Internet as far as I know. No realtime interaction or retrieval of info from the web. LM Studio has made many updates lately and added many new features so I can't say for certain. That is a great idea actually. I bet there is a python library that can be used to accomplish that.

@@videotronicmaker Thanks for the clarity. Many youtubers didn't talk abt this except you I think. I really wish they integrate this feature. It will be a game changer once and for all!!

Yes. Next will be Twiggy.

@@videotronicmaker 👍

Glad you enjoyed it

Yes. So have I. That is one goal. I hope to get there soon. For now GPT 3.5 is the closest option. Maybe GPT 4 o will be the next best. However, I haven't gotten the headset version working yet

@@videotronicmaker Yeah 4o looked pretty awesome, but Im sure it will be a while before its available to put on a home PC.. if its made public at all

I haven't tried it yet. I see some gguf Grok models on Hugging Face. Give it a shot. But I would not use a file that size on an SD card. I would use a file smaller than 15gb on an internal SSD drive otherwise you will have to wait 2 days to get a response to "Hello."

@@videotronicmaker external thunderbolt 4, newest version , with a throughput of 40Gb /sec, and a external NVME 20 GB/sec 2 TB ,..not an SD ...LOL

Typo? If you have a 334GB LLM then you'll need to run it on cloud hardware because otherwise you'd need 334GB of VRAM or RAM in order to run it locally. I tried a 60GB model and it was generating one token per minute. Not per second, per minute 😂

As far as I know, yes. If you're using software that incorporates the OpenAI library to run a local inference server, even if you're not directly using OpenAI's models, there could still be benefits. For example, the software might leverage the OpenAI library for compatibility, ease of integration, or additional functionalities provided by the library, such as optimized model loading, efficient inference processing, or support for various model architectures. In this scenario, using the OpenAI library within the software could streamline the setup and operation of the local inference server, enhancing its performance and capabilities. You might also want to read the GitHub repo info for the openai library as it contains more information regarding this. github.com/openai/openai-python And just to be clear, the openai library is not the same thing as the openai API or the chat GPT llm. The library contains all the other things required to use an llm. Everything is still local

Glad it was helpful!

Make sure you first create the file named system_message.txt. Second, make sure you put the correct path to that file in your code.

As far as I know there are two steps you can take to accomplish this. First, choose a model that speaks in your preferred language and then speak to the model in that language. Next, you can specify in the system prompt / system message that the model should always respond in a particular language. I have not attempted this yet because I do not speak any other language other than English however this would be my first approach if I wanted to do this.

Thanks for the comment. I tried that. For some reason it didn't work. I may try one more time. I definitely have that library installed in my env.

I'm glad to know I'm not the only one that was having trouble with that. I'm starting to think it was just me.

That was funny! Thanks for the laugh. Gemma said, "There is no evidence or record of a person setting foot on the moon." Well...at least it's consistent. 🤣

@@videotronicmaker yeah 😂, may be Google has begun to use data from an alternate universe to train these AI models. That's why Google's AI products are behaving so weirdly recently.

No matter what I change with the Hugging Face model hyperparameters via LM Studio it responds the same. So that is probably my error because when I go on Vertex AI, choose the Gemma model and use these settings on the right where it says, "Try out Gemma" : temperature = .8 top-p = .7 top-k = 30 ** Gemma responds: "Neil Armstrong was the first man to walk on the moon. So the answer is Neil Armstrong. I hope this helps! Please let me know if you have any further questions.**Best regards,** **The Answer Guy****Additional information:** Neil Armstrong was born in 1930 and was an American astronaut. He was the first human to set foot on the moon on July 20, 1969.He was a pioneer in the field of space exploration and his accomplishment is one of the greatest in human history. I hope this additional information is helpful.**" Here is the link to where I got that response: console.cloud.google.com/vertex-ai/publishers/google/model-garden/

@@videotronicmaker Thank you so much for the answer, I still haven't tried to use it on vertex AI, so I will try it now. Thanks!

Nope. I was wrong. I think I had the 7B selected in Vertex AI. I tried again to verify and no matter what, it said , "...no evidence..." For now, here is my conclusion. Google states: USE CASES Content Creation and Communication Text Generation: These models can be used to generate creative text formats such as poems, scripts, code, marketing copy and email drafts. Chatbots and Conversational AI: Power conversational interfaces for customer service, virtual assistants, or interactive applications. Text Summarization: Generate concise summaries of a text corpus, research papers, or reports. ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- I don't think the 2B models are trained on enough data to know the answer. The 7B does. I'm guessing that this is an example of the compromise when we choose a small model. I am making a video of my user experience now.

I definitely understand. The way I look at it is that it's not going to perform at the level of GPT 3.5 and 4 or at the level of Gemini Pro and Ultra however, out of all the small and fast models that you can find on hugging face I really believe this is the absolute best one. It has the ability to brainstorm and have conversations but it's definitely no GPT 3.5. I see it as a great way to get started on a local server if you have a small or low end computer. It's great for testing out and I believe that there is great potential for it to be trained and fine-tuned. I guess the short of it is that it may be the best starting point so far for models at that size. I would definitely be interested to hear some details of your experience because I am getting ready to make a part two to this where I go a little more in depth and take my time a little bit and make a more organized video. I made this video last night at 3:00 am and I was scrambling because it was a breaking news situation. Today I want to give it a little more time and use the q8 version.

It's coming out soon. Not released yet.

ElectroCookie has this one for $15. It's a 2 pack! I made a video about ElectroCookie shields for the Uno. They make good shields. This seems to be the best answer: www.amazon.com/ElectroCookie-Arduino-Prototype-Stackable-Expansion/dp/B0841ZFP1C/ref=sr_1_13?crid=CZFWT2PHTQUW&keywords=arduino+proto+shield&qid=1707607942&s=electronics&sprefix=arduino+proto+shield%2Celectronics%2C384&sr=1-13 This is the one I bought but it is not available right now: www.amazon.com/gp/product/B0169WHVGS/ref=ppx_yo_dt_b_search_asin_image?ie=UTF8&psc=1 Elegoo has this but it's a kit and costs $65: www.amazon.com/EL-KIT-008-Project-Complete-Ultimate-TUTORIAL/dp/B01EWNUUUA/ref=sr_1_1_sspa?crid=2W8V77GCU2ZX8&keywords=arduino+mega+kit&qid=1707607831&s=electronics&sprefix=arduino+mega%2Celectronics%2C386&sr=1-1-spons&ufe=app_do%3Aamzn1.fos.006c50ae-5d4c-4777-9bc0-4513d670b6bc&sp_csd=d2lkZ2V0TmFtZT1zcF9hdGY&psc=1

Shield without headers soldered on. store-usa.arduino.cc/products/arduino-mega-proto-shield-rev3-pcb

Glad it was helpful!

I appreciate that!

Glad to hear it!

Definitely. Coming shortly.

I found this video very helpful!!

Glad I could be of service ;-)

Here you go. Better natural language processing. Latest video with code: ruclips.net/video/lSnyTC3lh5o/видео.html

Yes. I saw your video. It appears that my shot starts just after yours ended. It was a very unique thing to see in the sky. It was unlike anything I ever saw before.

You are very welcome.