Gheorghe Chesler's picture

7 1 21

Gheorghe Chesler PRO

nightmedia

·

AI & ML interests

Exploring creative ways to improve LLM performance and make it fit

Recent Activity

liked a model about 2 hours ago

YOYO-AI/Qwen3-30B-A3B-Mixture-2507

updated a model about 20 hours ago

nightmedia/Qwen3-42B-A3B-2507-YOYO2-TOTAL-RECALL-Instruct-qx6-mxfp4-mlx

updated a collection about 23 hours ago

Runs On My 64GB Mac

View all activity

Organizations

None yet

New activity in nightmedia/unsloth-GLM-4.5-Air-qx64-mlx 2 days ago

Great Quant !

#1 opened 2 days ago by

New activity in nightmedia/GLM-4.5-Air-q5-hi-mlx 8 days ago

would you consider making a Q5 version with page size of 16?

#1 opened 26 days ago by deleted

New activity in DavidAU/Qwen3-Coder-42B-A3B-Instruct-TOTAL-RECALL-MASTER-CODER-M-768k-ctx 11 days ago

Any benchmarks to support improvements?

#1 opened 18 days ago by

New activity in nightmedia/gpt-oss-20b-q4-hi-mlx about 1 month ago

Any chance this would run on an iphone 16 pro?

#1 opened about 1 month ago by

New activity in nightmedia/Qwen3-30B-A3B-Thinking-2507-dwq4-mlx about 1 month ago

why are the weights larger than instruct model?

#1 opened about 1 month ago by

New activity in nightmedia/Qwen3-30B-A3B-Instruct-2507-q5-hi-mlx about 1 month ago

What is the meaning of “hi” in the model name?

#1 opened about 1 month ago by

New activity in mradermacher/model_requests 2 months ago

quantization request for Total-Recall v1.4

#1124 opened 2 months ago by