First trellis-coded LLM quantization decoder outside CUDA — an 8B model running its full forward pass in a browser on WebGPU. SOTA 3-bit quality (1.15x fp16).
-
Updated
Jul 24, 2026 - Python
#
trellis-coded-quantization
Here are 2 public repositories matching this topic...
Certified low-bit LLM weight quantization: rotation + derived-exponent equilibration + trellis-coded quantization with Hessian error feedback, a packed Triton runtime, and a KL/top-1/perplexity harness.
compression transformers pytorch quantization post-training-quantization llm gptq trellis-coded-quantization
-
Updated
Jul 11, 2026 - Python
Improve this page
Add a description, image, and links to the trellis-coded-quantization topic page so that developers can more easily learn about it.
Add this topic to your repo
To associate your repository with the trellis-coded-quantization topic, visit your repo's landing page and select "manage topics."