🧠 optimising digital brain

Sat Grewal PRO

satgeze

·

AI & ML interests

Optimise AI models for consumer grade systems resources

Recent Activity

liked a model 1 day ago

dealignai/DeepSeek-V4.1-Flash-UNCENSORED-FP8

liked a model 12 days ago

Mia-AiLab/Qwen3.8-27B-DFlash2-EXL3-5.0bpw

updated a model 12 days ago

satgeze/GLM-5.3-Flash-EXL3-TR3-3.5bpw

View all activity

Organizations

None yet

New activity in satgeze/Ornith-1.0-35B-1M-GGUF 23 days ago

Very good implementation of the 1M context, can we have same for Ornith-1.5-9B-GGUF, Ornith-1.5-35B-A3B-GGUF, Ornith-1.5-397B-GGUF (Q6_K)?

#2 opened 24 days ago by

New activity in satgeze/Qwen3.6-27B-DSpark about 2 months ago

Failed to launch with vLLM (latest version v.26.0)

#2 opened about 2 months ago by

~2.67x speedup in Mac with mlx-dspark

#1 opened about 2 months ago by

New activity in satgeze/Ornith-1.0-35B-1M-GGUF about 2 months ago

THIS IS BRILLIENT!!

#1 opened 2 months ago by

New activity in satgeze/Hy3-1M-GGUF about 2 months ago

Q6 gguf - "400 Bad Request"

#7 opened 2 months ago by

New activity in satgeze/Hy3-1M-GGUF 2 months ago

Radeon AI PRO R9700 IQ2_M

#6 opened 2 months ago by

dflash needed & speed is very slow at mtp

#4 opened 2 months ago by

Validated Q2_K on 3x 80GB A100 box - Very nice work!

#5 opened 2 months ago by

cudabenchmarktest

need higer quant

#1 opened 2 months ago by

Apple-silicon and CUDA datapoint with MTP on/off

#2 opened 2 months ago by

Q3_K_M - Results so far

#3 opened 2 months ago by

Apple-silicon and CUDA datapoint with MTP on/off

#2 opened 2 months ago by

need higer quant

#1 opened 2 months ago by

need higer quant

#1 opened 2 months ago by