Menu

#Ternary

5 posts

Feed·
5 of 5 posts
0Read More

TKintegers K5: 3.21× prefill and 3.10× decode throughput vs TorchAO FP8 on NVIDIA L4 [R] [P]

I built TKintegers, a balanced-ternary weight representation, converter, and GPU inference implementation. The result that stood out to me: the NVIDIA L4 has native FP8 tensor-core support, but my K5 implementation achieved higher throughput than the…

15s
Introducing 1-bit and Ternary Bonsai Image 4B: Image Generation for Local Devices
🖼️
0Read More

Introducing 1-bit and Ternary Bonsai Image 4B: Image Generation for Local Devices

Hacker News·6 months ago
#PFjrFPvc

Images generated from Ternary Bonsai Image 4BToday we’re releasing Bonsai Image 4B, a family of compact image-generation models designed to run high-quality diffusion inference on local hardware: from laptops to phones.Bonsai Image 4B comes in two…

15s
1-bit, 545 megabytes, zero API keys — local AI that beats GPT-5.4
🖼️
0Read More

1-bit, 545 megabytes, zero API keys — local AI that beats GPT-5.4

DEV Community·Vilius·6 months ago
#fMdkWkjh
#ai#llm#local#quantization#bonsai

1-bit, 545 megabytes, zero API keys — local AI that beats GPT-5.4 By Vilius Vystartas |...

15s