No abstract

No abstract

In this study, an optimum tempering temperature after a thermo-mechanical control process (TMCP) was proposed to improve the hydrogen-induced ductility loss of high-vanadium X80 pipeline steel. The results showed that with increasing tempering temperature from 450 to 650 °C, the size and quantity of granular bainite decreased but the spacing of deformed lath ferrite and the fraction of massive ferrite increased. The number of fine vanadium carbides increased as well. However, as the tempering temperature increased to 700 °C, the microstructure of T700 steel completely converted to massive ferrite and the grain size became larger. Additionally, the amount of nanoscale precipitates decreased again, and the mean size of precipitates evidently increased in T700 steel. The steel tempering at 650 °C, containing the most vanadium precipitates with a size less than 20 nm, had the lowest hydrogen diffusion coefficient and the best resistance to hydrogen-induced ductility loss.

No abstract

Neural machine translation (NMT) has achieved remarkable success in producing high-quality translations. However, current NMT systems suffer from a lack of reliability, as their outputs that are often affected by lexical or syntactic changes in inputs, resulting in large variations in quality. This limitation hinders the practicality and trustworthiness of NMT. A contributing factor to this problem is that NMT models trained with the one-to-one paradigm struggle to handle the source diversity phenomenon, where inputs with the same meaning can be expressed differently. In this work, we treat this problem as a bilevel optimization problem and present a consistency-aware meta-learning (CAML) framework derived from the model-agnostic meta-learning (MAML) algorithm to address it. Specifically, the NMT model with CAML (named CONMT) first learns a consistent meta representation of semantically equivalent sentences in the outer loop. Subsequently, a mapping from the meta representation to the output sentence is learned in the inner loop, allowing the NMT model to translate semantically equivalent sentences to the same target sentence. We conduct experiments on the NIST Chinese to English task, three WMT translation tasks, and the TED M2O task. The results demonstrate that CONMT effectively improves overall translation quality and reliably handles diverse inputs.