32 4 51

John Locke

johnlockejrr

[email protected]

AI & ML interests

NLP, OCR, AI

Recent Activity

liked a model 2 days ago

mittagessen/bytellama-10m-oscar

liked a model 2 days ago

rednote-hilab/dots.ocr

liked a Space 8 days ago

malper/unikud

View all activity

Organizations

liked 2 models 2 days ago

mittagessen/bytellama-10m-oscar

Updated 3 days ago • 51 • 1

rednote-hilab/dots.ocr

Image-Text-to-Text • 3B • Updated 3 days ago • 22.3k • 691

liked a Space 8 days ago

Unikud

🌖

liked a model 11 days ago

tau/tavbert-he

Fill-Mask • Updated Feb 8, 2022 • 305 • 3

updated a model 15 days ago

johnlockejrr/aramaic-diacritization-model

0.1B • Updated 15 days ago • 12

published a model 15 days ago

johnlockejrr/aramaic-diacritization-model

0.1B • Updated 15 days ago • 12

liked 3 Spaces 15 days ago

Hebrew-Yiddish Translator

📚

Masoretic Hebrew-Aramaic Translator

📚

Translate Masoretic Hebrew to Targumic Aramaic

Aramaic Targum Diacritization

📚

Add nikkud to Aramaic text

liked a Space 19 days ago

104

Qwen3 MT Demo

🏃

Translate text between multiple languages

reacted to mitkox's post with 🚀 21 days ago

Post

2086

I run Qwen3-Coder 480B locally on my Z8, with a 1-million token context window. It’s the equivalent of parallel-parking a Nimitz-class carrier in a kiddie pool. Thanks to whatever dark pact the llama.cpp, CUDA, and kernel folks signed, hybrid inferencing + VRAM↔RAM offload let me stream the model’s synapses across Xeon, RAM, and four lonely A6000s without summoning either the OOM killer or a small house fire.