🔄 In a Training Loop

Asankhaya Sharma

codelion

112 148 439

http://asankhaya.github.io/

AI & ML interests

Creator of OptiLLM, OpenEvolve, Adaptive Classifier, and Ellora. Pioneering a new category in AI infrastructure: inference-time compute for LLMs.

Recent Activity

updated a dataset about 2 hours ago

adaptive-classifier/ai-detector-data

new activity 2 days ago

litert-community/moonshine-tiny:Add measured Galaxy S26 NPU/GPU encoder rows

commentedon an article 2 days ago

Introducing @huggingface/kernels: 200+ WebGPU Kernels for Local AI

View all activity

Organizations

updated a dataset about 2 hours ago

adaptive-classifier/ai-detector-data

Viewer • Updated about 2 hours ago • 2.19k • 1.06k • 4

New activity in litert-community/moonshine-tiny 2 days ago

Add measured Galaxy S26 NPU/GPU encoder rows

#3 opened 7 days ago by

mlboydaisuke

commented on Introducing @huggingface/kernels: 200+ WebGPU Kernels for Local AI 2 days ago

@sroussey Agreed—the excluded 947 need a taxonomy before the 2.57× summary can travel. I'd want counts by operation and device for unsupported, ORT mismatch, new-kernel mismatch, and harness/timeout, with raw failing cases downloadable. Then Fleet can optimize only within a proven-correct compatibility envelope; fusion and per-device tuning become much safer.

New activity in zero-gpu-explorers/README 3 days ago

ZeroGPU clone fails during tensor packing with Errno 28 — 82.1 GB pack target

#185 opened 4 days ago by

Thananon

New activity in google/timesfm-3.0-pytorch 4 days ago

MCP server: call TimesFM-3 from Claude / Cursor (median + q10–q90)

👍❤️ 2

#2 opened 5 days ago by

DevairMd

liked a model 4 days ago

mlx-community/Qwen3.6-35B-A3B-OptiQ-4bit-REAP-19B

Text Generation • 19B • Updated 10 days ago • 658 • 1

commented on Introducing @huggingface/kernels: 200+ WebGPU Kernels for Local AI 4 days ago

The Fleet part may matter as much as the kernels. The 809 comparable cases on an M4 are useful, but browser, driver and device variance is where WebGPU tends to get painful. Will Fleet eventually expose per-device distributions, or is the plan to use the runs only for aggregate selection rules?