Modern natural language processing (NLP) research requires writing code. Ideally this code would provide a precise definition of the approach, easy repeatability of results, and a basis for extending the research. However, many research codebases bury high-level parameters under implementation details, are challenging to run and debug, and are difficult enough to extend that they are more likely to be rewritten. This paper describes AllenNLP, a library for applying deep learning methods to NLP research, which addresses these issues with easyto-use command-line tools, declarative configuration-driven experiments, and modular NLP abstractions. AllenNLP has already increased the rate of research experimentation and the sharing of NLP components at the Allen Institute for Artificial Intelligence, and we are working to have the same impact across the field.

Citrate may be used as a regional anticoagulant and the only buffering agent in CVVH with adequate treatment efficacy and safety. However, neither citrate nor heparin anticoagulation should be regarded as a therapeutic standard, since there is no advantage of one of these substances with regard to patient mortality.

We assessed clinical and duplex sonographic (CDS) findings, and outcome in patients with stenosis of the transplant renal artery (TRAS) or the aorto-iliac segment proximal to the graft (Prox-TRAS) treated with dilatation (PTA), stenting (PTAS) and surgery. Hypertension and graft dysfunction due to perfusion problems are rare. Clinical findings are nonspecific but CDS findings are helpful to select patients for angiography. Invasive treatment leads to clinical improvement. Surgery yields better results than PTA, but additional stenting will probably improve the outcome of angioplasty.

No abstract