Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

H3-metal – Native MiniMax-H3 inference for Apple Silicon

github.com

Post reply on HN
Print view View original

31–40 of 108 posts

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#31

Post by yieldcrv »

alexgoodhart wrote: post #20
Meleagris wrote: post #6

I've been using MiniMax H3 on my M5 Pro 64GB MacBook Pro through ComfyUI. It works extremely well. I had to modify the default ComfyUI workflows to use a GGUF quant (city96's ComfyUI-GGUF custom node, UnetLoaderGGUF in place of the stock loader) [0]. I use the model labeled Q5_K_M. There is Q8_0 available as well, which is 34GB and fits fine in 64GB unified memory if you keep resolution modest. The main issue is spee…

I wonder how much faster your m5 pro is compared to my M1 Max @ 64gb

wait till the M7 bro you’re almost there, rumor has it that Apple is skipping the M6 but it still might be 2028

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#33

Post by SwellJoe »

robotresearcher wrote: post #25
abhinai wrote: post #4

How similar are Jeff Dean and Salvatore Sanfilippo?

People are really good at stuff. I noticed on a bar TV the other day that some of the Chromecast screensaver landscape photo credits were to Peter Norvig. They were really lovely pictures.

I've run into Peter Norvig twice. Once at a YC event; the other when I parked my motorhome in front of his house in Palo Alto for a couple of days while visiting a friend who happened to live on the same street (not on purpose, I didn't know it was his house, it was just where I found sufficient open street parking for a huge motorhome, big houses with fewer cars on the street than on my friend's block). I ran into him while walking my dog, he asked about the motorhome and we talked travel. He was lovely both times. Not everyone is nice about a big motorhome parking on their block, especially in California, but he was friendly.

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#34

Post by moshun »

azinman2 wrote: post #17
behnamoh wrote: post #12

Earlier quoted context omitted.

when you have enough money to not have to worry about anything, you can go back to your hobbies. in this case, his hobby is programming.

Being a world class talent is independent of financial situation.

“I am, somehow, less interested in the weight and convolutions of Einstein's brain than in the near certainty that people of equal talent have lived and died in cotton fields and sweatshops."— Stephen Jay Gould

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#35

Post by antirez »

Meleagris wrote: post #6

I've been using MiniMax H3 on my M5 Pro 64GB MacBook Pro through ComfyUI. It works extremely well. I had to modify the default ComfyUI workflows to use a GGUF quant (city96's ComfyUI-GGUF custom node, UnetLoaderGGUF in place of the stock loader) [0]. I use the model labeled Q5_K_M. There is Q8_0 available as well, which is 34GB and fits fine in 64GB unified memory if you keep resolution modest. The main issue is spee…

This implementation is much faster on my M5 Max, like a few minutes for the same video, but on an M5 Max with 128GB, didn't test on M5 Pro. About memory, could be executed on 64GB with a few changes.

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#36

Post by Manfrednotfunny »

antirez wrote: post #35
Meleagris wrote: post #6

I've been using MiniMax H3 on my M5 Pro 64GB MacBook Pro through ComfyUI. It works extremely well. I had to modify the default ComfyUI workflows to use a GGUF quant (city96's ComfyUI-GGUF custom node, UnetLoaderGGUF in place of the stock loader) [0]. I use the model labeled Q5_K_M. There is Q8_0 available as well, which is 34GB and fits fine in 64GB unified memory if you keep resolution modest. The main issue is spee…

This implementation is much faster on my M5 Max, like a few minutes for the same video, but on an M5 Max with 128GB, didn't test on M5 Pro. About memory, could be executed on 64GB with a few changes.

Memory bandwidtih between pro and max is double. 300gb/s vs. 600gb/s btw.

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#37

Post by hnfong »

onionisafruit wrote: post #5
abhinai wrote: post #4

How similar are Jeff Dean and Salvatore Sanfilippo?

My favorite Jeff Dean fact is that he’s also antirez. Which reminds me of my favorite Salvatore Sanfilippo fact. He’s also Jeff Dean

The claims seem fitting from a user named "onionisafruit"...

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#38

Post by Myzura »

Meleagris wrote: post #6

I've been using MiniMax H3 on my M5 Pro 64GB MacBook Pro through ComfyUI. It works extremely well. I had to modify the default ComfyUI workflows to use a GGUF quant (city96's ComfyUI-GGUF custom node, UnetLoaderGGUF in place of the stock loader) [0]. I use the model labeled Q5_K_M. There is Q8_0 available as well, which is 34GB and fits fine in 64GB unified memory if you keep resolution modest. The main issue is spee…

How much free space do you have left after running this llm model? Have you tried to develop your own model with the M5?

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#39

Post by Havoc »

yieldcrv wrote: post #32

Alright I’ve been afraid to ask but have been having trouble finding What are some adult entertainment workflows in comfyui, I need best loras, best prompts to start with and the communities, are they on telegram or something?

There will Reddit subs for it though couldn’t tell you which off top of my head

I’d personally steer clear of messaging platforms for this - who knows what one might stumble into there

Re: H3-metal – Native MiniMax-H3 inference for Apple Silicon

#40

Post by antirez »

Manfrednotfunny wrote: post #36
antirez wrote: post #35

Earlier quoted context omitted.

This implementation is much faster on my M5 Max, like a few minutes for the same video, but on an M5 Max with 128GB, didn't test on M5 Pro. About memory, could be executed on 64GB with a few changes.

Memory bandwidtih between pro and max is double. 300gb/s vs. 600gb/s btw.

Does not matter much in this case. GPU bound.
Post reply on HN