Harry

fun haver

[Harry @Harshitpal0844]

@max_spero_ u d real ABG

[Harry @Harshitpal0844]

@0xSero hello

[Harry @Harshitpal0844]

@CathPoaster no thanks, im good

[Harry @Harshitpal0844]

@usr_bin_roygbiv valid

[Harry @Harshitpal0844]

@victoriazmei dawg u r asian how can you not end up in tech

[Harry @Harshitpal0844]

@Kenneth_Marino why did you leave? hows utah? i graduated from utah in may 2025 with MS from Tolgas lab ( worked on multimodal interpretability and medical imaging). would you say is this suitable time for getting a phd or being in academia in general for AI?

[Harry @Harshitpal0844]

@mil000 tf we don’t

[Harry @Harshitpal0844]

@mil000 very valid take

[Harry @Harshitpal0844]

@Brennan_Lup nah ima gatekeep my place

[Harry @Harshitpal0844]

@ivanfioravanti solid start

[Harry @Harshitpal0844]

@ivanfioravanti keep us posted!!

[Harry @Harshitpal0844]

@ivanfioravanti hey Ivan, absolutely love your work. I wanted to whats the best coding model for 256GB M3 ultra? Running ds4 lately though Pi; but wanna try newer models (k2.7, glm5.2, step) specifically for coding

[Harry @Harshitpal0844]

@oscmansan me

[Harry @Harshitpal0844]

@JustinLin610 retarded model

[Harry @Harshitpal0844]

@stash_pomichter no thanks

[Harry @Harshitpal0844]

@Prince_Canuma @GoogleDeepMind it was stuck on infinite thinking loop with outputtin trash; still debugging on my end, will file a issue on gh if cant resolve

[Harry @Harshitpal0844]

@Prince_Canuma 🐐🐐🐐🐐

[Harry @Harshitpal0844]

@mudler_it @ggml_org @NVIDIAAIDev @LocalAI_API any stats in how does it compare to coreml??

[Harry @Harshitpal0844]

@0xSero @ivanfioravanti I tried minicpm 1b and its just horrible like I couldn’t even get it to answer basic questions properly; lfm 2.5 6bit is perfect for my needs