DeepSeek and Qwen are truly amazing!

gotnull said:
thank you for sharing the name of the projec
Click to expand...
I still have it in my TODO list. Recently I bought OrangePI 5 max that has rk3588 SOC in it. It's far from ideal though how rockchip (and others along the chain) do the support. It's half baked, half glued software with blob bootcode from China. Even u-boot relies on rockchip's blob SPL. I treat it with certain level of scepticism. And yes, I know even RPI is not opened in this regard.
It's powerfull little thingy though.

Starting point for this project in my case: rknn. But basically I'd need to google how to setup local LLM with this chipset. There are some YT videos for this board where people tested chat bots on this.
 
gotnull said:
That's probably the best usage of an AI I've read so far, it gives hope, thank you for sharing the name of the project because I never heard of it until today.
Google being Google let's hope they won't abandon it once the hype around AI will fade.
Click to expand...

Yeah, there's Google's "oh, squirrel" mentality.

I hope somebody hacks up DeepSeek so that it can do the same thing as notebooklm, self-hosted.That would give future-proofing.
 
_martin said:
I looked over my over the 20 years worth of pdfs I collected. I don't want to delete them but frankly I never check them out. If I need something I google it.
Click to expand...
Same here.

When I want to study something, I learned long ago to search through my sorted directories of these things from the terminal. I find things that are out of date or not what I thought they were. So I delete them. Often I find the solution to what I'm looking for. Using a graphical interface, like a desktop thing, blurs it all and I don't do anything.

I use find and grep a lot.
 
drhowarddrfine said:
I use find and grep a lot.
Click to expand...
I agree, I do the same. I have my private/notes dir with all the various notes (plain files) I created over all these years. Grepping around that is easy and does the job perfectly. The day I learned of grep's -H my life got easier too. ;)

And while you could also do pdf2text and grep that (doesn't solve the issue with the pdf full of pics) it's just easier then to google/search online. At least that's what I keep doing.
If (to be on safe side not saying when) I do play with these models I will share my experience.
 
_martin said:
I still have it in my TODO list. Recently I bought OrangePI 5 max that has rk3588 SOC in it. It's far from ideal though how rockchip (and others along the chain) do the support. It's half baked, half glued software with blob bootcode from China. Even u-boot relies on rockchip's blob SPL. I treat it with certain level of scepticism. And yes, I know even RPI is not opened in this regard.
It's powerfull little thingy though.

Starting point for this project in my case: rknn. But basically I'd need to google how to setup local LLM with this chipset. There are some YT videos for this board where people tested chat bots on this.
Click to expand...

I'm building my own tablet based on FreeBSD and equipped with the RockPro RK399,but i see that your rkmn tool does not support it,why ?
 
cracauer@ said:
Yeah, there's Google's "oh, squirrel" mentality.

I hope somebody hacks up DeepSeek so that it can do the same thing as notebooklm, self-hosted.That would give future-proofing.
Click to expand...

Hello Master. I read the DeepSeek requires a very low amount of GPUs to train a dataset. Has it reached the point where it can be done using only a gpu like the RTX 2080 ti ? or maybe a 4090 or only a couple of them ? My dream is always the same. To train a dataset with the raw commands used by bhyve to create an AI that can do what we want starting by a well formatted prompt. I think this is the dream of a lot of virtualization lovers.
 
ZioMario said:
Hello Master. I read the DeepSeek requires a very low amount of GPUs to train a dataset. Has it reached the point where it can be done using only a gpu like the RTX 2080 ti ? or maybe a 4090 or only a couple of them ? My dream is always the same. To train a dataset with the raw commands used by bhyve to create an AI that can do what we want starting by a well formatted prompt. I think this is the dream of a lot of virtualization lovers.
Click to expand...

No. It still uses datacenters full of GPUs for training. Just less so. When you install it locally you download it along with the trained model.

Apart from the computational load you would also have to download most text on the Internet to train it yourself.

The interesting part of playing with it isn't the training, though. You can use and analyze it just running the inference locally. A beefy NVidia GPU is enough for that, you don't need a 4090.
 
Hi folks, I am looking for a graphic card to run Deepseek on my home desktop (i5-12400 + 64G DDR5, FreeBSD 14.2). As you can see, it is a budget build so the graphic card should not be too modern as well.

What are the general recommendations? Would an intel ARC work?
 
I5 ? Intel ARC ? That's a low budget hardware...
Intel ARC may have some problem with FreeBSD and anyway,it does not have a good hardware acceleration..
 
I've used online chatgpt and deepsink (which requires a sign in but I have a throwaway address to use) a few times. There have been one or two (or more perhaps) when I got better results than I did with a web search. On the other hand, it didn't help me get sound on a Beelink SER5, I got that from a web search, though it took me much longer, nor did I try every suggestion given me by the LLMs.

I'm not a coder nor developer, but it seems as if it's being greatly overhyped--it seems like it just a large database but everyone calls it AI. (I have, and I'm sure many of you, have various non-complimetary substitutions for the initials.
 
DeepSeek vs. Copilot - which one do you thinks has better AI ?

Q: What is bhyve virtual machine?

A: DeepSeek

A BHYVE virtual machine (VM) is a type of virtual server that allows you to create isolated environments for
running applications, services, and databases on top of physical hardware. The term "BHYVE" may refer to a
specific vendor or platform, but generally, virtual machines are used across various cloud providers like AWS,
Azure, Google Cloud, and others.

### Key Features of BHYVE Virtual Machines:
1. **Isolation**: Each VM operates independently, allowing multiple applications or services to run simultaneously
without interfering with each other.
2. **Scalability**: You can scale up or down resources (CPU, memory, storage) based on demand.
3. **Portability**: VMs can be easily transferred between physical machines.
4. **Resource Efficiency**: Resources like CPU and memory are allocated per VM, allowing for better utilization.

### Benefits of Using Virtual Machines:
1. **Cost-Effectiveness**: Leverage existing hardware resources without high capital investment.
2. **Improved Performance**: Dedicated resources lead to faster performance for each application.
3. **Flexibility**: Test and deploy applications in different environments (development, testing, production) on
the same infrastructure.

### Example Use Cases:
- Running multiple development environments (e.g., Node.js, Python).
- Hosting virtual servers for web applications or databases.
- Isolating containerized applications (e.g., Docker).


A: Copilot

bhyve (pronounced "bee-hive") is a hypervisor/virtual machine manager for FreeBSD. It allows you to run multiple virtual machines (VMs) on a single physical machine1. Here are some key features:

Guest OS Support: Supports a wide range of guest operating systems, including various versions of Windows (Vista, 7, 8/8.1/8.2, 10, 11), and Windows Server versions.

Hardware Support: Works on Intel and AMD processors that support the POPCNT feature, and experimentally on ARM64/aarch64 processors.

Storage and Networking: Supports multiple storage and networking back-ends, including NVMe emulation, AHCI devices, and VirtIO drivers.

Boot Options: Supports UEFI, FreeBSD loader, and GRUB booting.

Advanced Features: Includes PCI Pass-Through, integrated VNC and pfs servers, and TPM passthrough/emulation2.

It's a powerful tool for anyone looking to run multiple operating systems on a single host machine. Does this help clarify things for you?

1) wiki.freebsd.org
2) wiki.freebsd.org
 
After plying around with DeepSeek, in my opinion, it's useless for specific or even general information gathering.
I don't want to pollute this thread with more nonsense from DeepSeek and its answer on "How to install bhyve in FreeBSD?"
 
it seems as if it's being greatly overhyped--it seems like it just a large database but everyone calls it AI
Click to expand...

thats pretty much the conclusion i came to as well
its just a offline database you can query

also what it knows is tied to when the "database" was created
so obviously the model doesnt know about anything after that date it was created
 
NapoleonWils0n said:
thats pretty much the conclusion i came to as well
its just a offline database you can query

also what it knows is tied to when the "database" was created
so obviously the model doesnt know about anything after that date it was created
Click to expand...

I agree. But it also seems to me that it has some notable gaps in the knowledge. I asked it which Italian soccer team has the woodpecker emblem and each time it answered with a different team. It never got it right. The answer was Ascoli, the team I support. ChatGPT replied correctly the first time. Maybe it works better than ChatGPT in some very specific scenario that we haven't tested ?
 
Keep in mind that DeepSeek is open source. Because of that it has taken university courses and research by storm.

This might lead to an accelerated rate of fixes either in the original code or an American fork of it.
 
You must log in or register to reply here.
Back
Top