126 79 1977

Nicky

NickyNicky

AI & ML interests

None yet

Recent Activity

liked a dataset 2 days ago

mlabonne/natural_reasoning-formatted

liked a model 2 days ago

google/siglip2-base-patch16-512

liked a Space 3 days ago

nanotron/ultrascale-playbook

View all activity

Organizations

NickyNicky's activity

liked a dataset 2 days ago

mlabonne/natural_reasoning-formatted

Viewer • Updated 2 days ago • 1.15M • 56 • 7

liked a model 2 days ago

google/siglip2-base-patch16-512

Zero-Shot Image Classification • Updated 3 days ago • 2.65k • 3

liked a Space 3 days ago

1.36k

The Ultra-Scale Playbook

🌌

The ultimate guide to training LLM on large GPU Clusters

liked a model 4 days ago

stepfun-ai/stepvideo-t2v

Text-to-Video • Updated 5 days ago • 1.04k • 311

liked a dataset 4 days ago

facebook/natural_reasoning

Viewer • Updated 2 days ago • 1.15M • 1.24k • 162

upvoted a paper 4 days ago

Magma: A Foundation Model for Multimodal AI Agents

Paper • 2502.13130 • Published 5 days ago • 42

liked a model 5 days ago

agents-course/notebooks

Updated 5 days ago • 106

liked a dataset 5 days ago

bethgelab/CuratedThoughts

Viewer • Updated 6 days ago • 245k • 158 • 26

liked a model 6 days ago

stepfun-ai/Step-Audio-Chat

Audio-Text-to-Text • Updated 6 days ago • 713 • 358

reacted to Jaward's post with 🔥 6 days ago

Post

3782

Finally here it is: a faster, custom, scalable GRPO trainer for smaller models with < 500M params, can train on 8gb ram cpu, also supports gpu for sanity sake (includes support for vllm + flash attention). Using smolLM2-135M/360M-instructs as ref & base models. Experience your own “aha” moment 🐳 on 8gb ram.
Code: https://github.com/Jaykef/ai-algorithms/blob/main/smollm2_360M_135M_grpo_gsm8k.ipynb