Quantize by Drift: Label-Free Mixed-Precision Post-Training Quantization for Text Embedders

A self-hosted assistant is too expensive because its model uses most of the GPU memory.

What you would change

Try the paper's supported lower-precision weights on a copy of your serving setup.

Try the paper's supported lower-precision weights on a copy of your serving setup.

Compare against

The same model at its original precision, on the same GPU.

1. Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.
2. Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.
3. Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.
4. Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test.

What to measure

Peak memory, task pass rate, latency, and total serving cost at the same context and load.

When to keep it

Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation.

recommendation

test

research area

quantization

technical change

Label-free drift-based mixed-precision PTQ for embedders

published

2026-10-06

The label-free sensitivity signal is cheap and reproducible on constrained infrastructure, but the authors' own sealed evaluation shows drift-based allocation loses to uniform quantization at the main budget, so it merits workload-specific validation at aggressive budgets rather than direct adoption.

1. claim 01Paperraft inference
   Read the brief interpretation

   The method replaces relevance-label-based or search-based per-module sensitivity estimation for mixed-precision post-training quantization of text embedders with a label-free signal: quantize one module, re-encode the corpus, and measure how far output embeddings drift, then allocate precision additively under a packed-byte budget. It costs one quantization pass per module plus a full corpus re-encoding per measurement, with no labels, no search, and only modest engineering complexity. It can fail as an allocation objective: at the main budget it underperformed same-budget uniform precision on all three sealed embedders, so it is justified mainly at stressed budgets where uniform precision collapses, and fine-grained redistribution around a strong uniform operating point remains unresolved.

   This is a Paperraft inference. It is not pinned to a verified PDF passage.

   Inspect the source
2. claim 02Paperraft inference

   Configuration-level drift orders mixed-precision plans against held-out retrieval quality at macro Spearman 0.911 across five embedders; on three sealed held-out embedders it beats the prior LieQ criterion 3/3 at the main budget but is numerically below same-budget uniform precision on all three (-0.99, -0.85, -1.01 points).

   This is a Paperraft inference. It is not pinned to a verified PDF passage.

   Inspect the source