Yassine Ennaour's picture

Yassine Ennaour

Lyte

·

AI & ML interests

None yet

Recent Activity

upvoted a paper about 8 hours ago

Rewarding the Rare: Uniqueness-Aware RL for Creative Problem Solving in LLMs

reacted to danielhanchen's post with ❤️ about 9 hours ago

You can now do reinforcement learning training with 7× longer context and no accuracy loss, via our new batching algorithms. Long reasoning chains in RL are costly, but now we enable you to train gpt-oss with GRPO & reach 380K context on a 192GB GPU. Blog: https://unsloth.ai/docs/new/grpo-long-context

reacted to danielhanchen's post with 🚀 about 9 hours ago

You can now do reinforcement learning training with 7× longer context and no accuracy loss, via our new batching algorithms. Long reasoning chains in RL are costly, but now we enable you to train gpt-oss with GRPO & reach 380K context on a 192GB GPU. Blog: https://unsloth.ai/docs/new/grpo-long-context

View all activity

Organizations

Lyte 's models 8

Lyte/Gemma-3-1B-Moroccan-Instruct

Text Generation • 1.0B • Updated Sep 11, 2025 • 51 • 1

Lyte/Whisper-Small-Darija

Automatic Speech Recognition • 0.2B • Updated Aug 27, 2025 • 1

Lyte/whisper-base-darija

72.6M • Updated Mar 27, 2025 • 6

Lyte/Whisper-Tiny-Darija

37.8M • Updated Mar 26, 2025 • 4

Lyte/QuadConnect2.5-0.5B-v0.1.1b

Text Generation • 0.5B • Updated Mar 1, 2025 • 4

Lyte/QuadConnect2.5-1.5B-v0.1.0b

Text Generation • 2B • Updated Feb 28, 2025 • 89 • 1

Lyte/QuadConnect2.5-0.5B-v0.0.9b

Text Generation • 0.5B • Updated Feb 27, 2025 • 38

Lyte/Llama-3.2-3B-Overthinker

Text Generation • 3B • Updated Oct 20, 2024 • 57 • 21