Serving systems choose a weight format for each model and deployment. A low-cost screen predicts the resulting KL divergence from the unquantized model's Fisher information using $\tfrac12Δ^\top FΔ$, the expansion underlying Hessian-aware and second-order post-training quantizers.
Experimental Evaluation
We assess the reliability of this screen in GPU batches with channel-wise int8, per-tensor fp8 and group-wise int4 on:
- Qwen3-8B
- Qwen3-4B
- OLMo-2-7B
- Qwen3-32B
- Qwen3-30B-A3B
Relative predictions hold across a 16-fold range of perturbation scales, int4 group sizes and rounding modes, while absolute predictions are heavy-tailed. A trimmed average brings 10 of 12 cells on Qwen3-4B and OLMo-2 within 15% of measured KL. The MoE model exhibits nonquadratic response at small perturbations, with int8 KL scaling as $α^{0.63}$.
Error Tracing and Resolution
Computing per-sequence gradients in bf16 can cause fourfold prediction errors. We trace these errors to arithmetic in the lower half of the backward pass and eliminate them by using fp32 there. Batch-invariant kernels run at 0.55 times the throughput, while changes in reduction order introduce a floor of 0.17-0.22 nats per sequence.
System Integration and Results
We build a screener with one streamed fp32 Fisher pass, trimmed tensor-block predictions with a band, a flag for aligned tensors and per-layer format allocation. On the unseen Mistral-7B-v0.3 family, all six predictions fall within the band, all three KL budgets are met, and vLLM throughput is predicted within 0.4%.