Search Results for "budgets"

FastViT is an efficient vision backbone family that blends convolutional inductive biases with transformer capacity to deliver strong accuracy at mobile and real-time inference budgets. Its design pursues a favorable latency-accuracy Pareto curve, targeting edge devices and server scenarios where throughput and tail latency matter. The models use lightweight attention and carefully engineered blocks to minimize token mixing costs while preserving representation power. Training and inference recipes highlight straightforward integration into common vision tasks such as classification, detection, and segmentation. ...

SlowFast is a video understanding framework that captures both spatial semantics and temporal dynamics efficiently by processing video frames at two different temporal resolutions. The slow pathway encodes semantic context by sampling frames sparsely, while the fast pathway captures motion and fine temporal cues by operating on densely sampled frames with fewer channels. Together, these two pathways complement each other, allowing the network to model both appearance and motion without...

...The GLM-Z1-32B-0414 line adds deeper mathematical, coding, and logical reasoning via extended reinforcement learning and pairwise ranking feedback, while GLM-Z1-Rumination-32B-0414 introduces a “rumination” mode that performs longer, tool-using deep research for complex, open-ended tasks. A lightweight GLM-Z1-9B-0414 brings many of these techniques to a smaller model, targeting strong reasoning under tight resource budgets.

MiniMax-M1 is presented as the world’s first open-weight, large-scale hybrid-attention reasoning model, designed to push the frontier of long-context, tool-using, and deeply “thinking” language models. It is built on the MiniMax-Text-01 foundation and keeps the same massive parameter budget, but reworks the attention and training setup for better reasoning and test-time compute scaling. Architecturally, it combines Mixture-of-Experts layers with lightning attention, enabling the model to...

...Training ran on TPU v6e using JAX and Pathways with privacy-preserving algorithms (DP-SGD, truncated Poisson subsampling) and DP scaling laws to balance compute and privacy budgets. Benchmarks on the 1B pre-trained checkpoint show expected utility trade-offs (e.g., HellaSwag 10-shot 39.09, BoolQ 0-shot 62.04, PIQA 0-shot 68.00), reflecting its privacy-first design.