[QaboosbinSaidAlSaid]

[*]

https://truleap.net/wp-content/uploads/2024/01/Jan-Week-1.jpg

Hidden Content

You must register or login to view this content.

[DredgenSun]

Personally I still use DeepSeek with an offline LLM platform, don't totally trust online LLM's anymore...

[QaboosbinSaidAlSaid]

Personally I still use DeepSeek with an offline LLM platform, don't totally trust online LLM's anymore...

thats amazing on what system configs u use it ? what deepseek model

[DredgenSun]

Personally I still use DeepSeek with an offline LLM platform, don't totally trust online LLM's anymore...

thats amazing on what system configs u use it ? what deepseek model

https://github.com/deepseek-ai/DeepSeek-V3

This running though ;

https://lmstudio.ai/ 

This


I originally had a QWEN node running QWENV3-14b LLM on my offline Video Diffusion workflow, as it adheres to better contextualization & prompt enhancement for producing better output from using Wan2.2 Video checkpoints.


I'm thinking WAN have given up with Open Source, as we're already on Wan2.6 and Wan2.5 has been out for awhile, and that's a fabulous video model imho

[exxxiko]

Impossible to use it...

This forum account is currently banned. Ban Length: Permanent (N/A Remaining)
Ban Reason: Contact Administration.

[QaboosbinSaidAlSaid]

Personally I still use DeepSeek with an offline LLM platform, don't totally trust online LLM's anymore...

thats amazing on what system configs u use it ? what deepseek model

https://github.com/deepseek-ai/DeepSeek-V3

This running though ;

https://lmstudio.ai/ 

This


I originally had a QWEN node running QWENV3-14b LLM on my offline Video Diffusion workflow, as it adheres to better contextualization & prompt enhancement for producing better output from using Wan2.2 Video checkpoints.


I'm thinking WAN have given up with Open Source, as we're already on Wan2.6 and Wan2.5 has been out for awhile, and that's a fabulous video model imho

YEAH WAN IS NUTS CHINESE DEVS WORKING WILD ON THAT SHIT ITS FIRE 

DEEPSEEK IS THE BEST QWEN IS GOAT

[rikaso]

Personally I still use DeepSeek with an offline LLM platform, don't totally trust online LLM's anymore...

thats amazing on what system configs u use it ? what deepseek model

https://github.com/deepseek-ai/DeepSeek-V3

This running though ;

https://lmstudio.ai/ 

This


I originally had a QWEN node running QWENV3-14b LLM on my offline Video Diffusion workflow, as it adheres to better contextualization & prompt enhancement for producing better output from using Wan2.2 Video checkpoints.


I'm thinking WAN have given up with Open Source, as we're already on Wan2.6 and Wan2.5 has been out for awhile, and that's a fabulous video model imho

I would be very interest to know about that, tutorial, the usage and if there are no limitation.
Could be interesting to have it's own LLM locally, but should imply so many resource and assume would not be efficient. (but at which level compared to those given in cloud)

This forum account is currently banned. Ban Length: Permanent (N/A Remaining)
Ban Reason: Contact Administration.

[AnarchyBin]

Personally I still use DeepSeek with an offline LLM platform, don't totally trust online LLM's anymore...

thats amazing on what system configs u use it ? what deepseek model

https://github.com/deepseek-ai/DeepSeek-V3

This running though ;

https://lmstudio.ai/ 

This


I originally had a QWEN node running QWENV3-14b LLM on my offline Video Diffusion workflow, as it adheres to better contextualization & prompt enhancement for producing better output from using Wan2.2 Video checkpoints.


I'm thinking WAN have given up with Open Source, as we're already on Wan2.6 and Wan2.5 has been out for awhile, and that's a fabulous video model imho

what kind of specs do you have because those are big models imo

This forum account is currently banned. Ban Length: Permanent (N/A Remaining)
Ban Reason: Scraping. If you feel this is incorrect: http://breached4wtyw5fb45zj7sggnoazgv3aohme2zftkrndhvo76d5q5uad.onion/

[luxyfoxy]

Personally I still use DeepSeek with an offline LLM platform, don't totally trust online LLM's anymore...

thats amazing on what system configs u use it ? what deepseek model

https://github.com/deepseek-ai/DeepSeek-V3

This running though ;

https://lmstudio.ai/ 

This


I originally had a QWEN node running QWENV3-14b LLM on my offline Video Diffusion workflow, as it adheres to better contextualization & prompt enhancement for producing better output from using Wan2.2 Video checkpoints.


I'm thinking WAN have given up with Open Source, as we're already on Wan2.6 and Wan2.5 has been out for awhile, and that's a fabulous video model imho

I would be very interest to know about that, tutorial, the usage and if there are no limitation.
Could be interesting to have it's own LLM locally, but should imply so many resource and assume would not be efficient. (but at which level compared to those given in cloud)

it for sure requires a lot of power to run that I guess

[QaboosbinSaidAlSaid]

Impossible to use it...

for real on my potato pc