Archsloth fixes two AutoRound flags, says Qwen3-4B drifts less
The project reports 20.9% to 54.4% lower KL divergence at the same 2.50 GB size, with calibration text doing more work than bit allocation.
By RuntimeWire Staff · Published
Primary source: Hugging Face Newsroom
Why it matters
Local-model distributors can ship files with the same size and format that preserve materially different behavior. Quantization recipes and calibration data deserve the same scrutiny as model weights and benchmarks.
The contributors behind Archsloth published a correction on September 14th to a quantization workflow they had used for months, saying two changed AutoRound settings produced a version of Alibaba's Qwen3-4B that stayed closer to the original model without adding a single byte.…
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.