Mostafa Elhoushi's picture

34 5

Mostafa Elhoushi

melhoushi

·

AI & ML interests

Make ML faster, smaller, smarter.

Recent Activity

commented on a paper 18 days ago

Skip a Layer or Loop it? Test-Time Depth Adaptation of Pretrained LLMs

upvoted a paper 18 days ago

Skip a Layer or Loop it? Test-Time Depth Adaptation of Pretrained LLMs

authored a paper 22 days ago

CHAI: Clustered Head Attention for Efficient LLM Inference

View all activity

Organizations

Articles 1

Article

59

Faster Text Generation with Self-Speculative Decoding

Papers 11

arxiv:2507.04610

arxiv:2506.00204

arxiv:2505.20309

arxiv:2410.00215

models 23

melhoushi/layerskip-Llama-2-7B-top_v2

melhoushi/layerskip-Llama-3.2-1B-top_v2

1B • Updated Mar 17 • 1

melhoushi/layerskip-Llama-2-7b-hf-top_v2

melhoushi/layerskip-SmolLM2-135M-top_v2

0.1B • Updated Mar 16 • 4

melhoushi/layerskip-huggingface-smollm2-135m-topv1

0.1B • Updated Mar 16 • 7

melhoushi/layerskip-huggingface-smollm-135m-topv1

0.1B • Updated Mar 16 • 3

melhoushi/layerskip-llama3.2-1b-topv1

1B • Updated Mar 16 • 2

melhoushi/layerskip-llama2-7b-topv1

melhoushi/exitskip-llama2-7b-topv1-v1

7B • Updated Jan 24 • 5

melhoushi/layerskip-llama2-7b-topv1-v4

7B • Updated Jan 23 • 6

datasets 0

None public yet