# Adaptive compute techniques yield significant inference speedups across models > Source: > Published: 2026-08-22 05:00:00+00:00 FlashMorph slashes the cost of designing hybrid attention models, needing only 20 M tokens and 2.1 GPU‑hours for layer selection. “FlashMorph uses only 20M tokens for hybrid layer selection, requiring FLOPs and 2.1 GPU hours.” [[1]](https://arxiv.org/abs/2606.30562) This represents a substantial reduction compared with prior heuristic searches that sweep millions of configurations on full datasets, as reported by the authors. The method also preserves retrieval quality at short and medium context lengths, matching the performance of dense baselines. “FlashMorph achieves near‑perfect accuracy on NIAH‑Single‑1 and delivers strong performance on the more challenging NIAH‑Single‑2 and NIAH‑Single‑3 settings, particularly at short and medium context lengths.” [[1]](https://arxiv.org/abs/2606.30562) Even with a budget‑constrained mix of full and linear attention layers, recall drops are negligible. Before FlashMorph, hybrid conversion relied on fixed patterns or per‑layer scores that ignored interactions between layers, forcing engineers to hand‑craft budgets or run expensive grid searches. Those heuristics treated each layer’s importance in isolation, leading to suboptimal trade‑offs between speed and long‑context fidelity. FlashMorph frames layer selection as a global subset‑optimization problem, freezing the pretrained weights while jointly learning binary gates on synthetic retrieval data. Linearization regularisation pushes the model toward linear attention wherever possible, after which discrete gates instantiate the final architecture followed by standard distillation. The paper’s evaluation is limited to long‑context retrieval benchmarks; it does not test generation quality in LLMs, diffusion sampling speed, or text‑to‑image pipelines. Moreover, the synthetic data used for gate training may not capture distribution shifts encountered in downstream applications. This suggests that extending budget‑constrained morphing to generative tasks will require task‑specific proxy datasets and possibly fine‑tuning of the linear branches. If hybrid conversion can be automated with such low overhead, engineers should replace handcrafted layer patterns with a FlashMorph‑style search as the default deployment step for any model that must handle extended contexts. The resulting hybrids promise near‑baseline accuracy while cutting memory use and latency enough to enable real‑time inference on commodity hardware.