Adapters are light-weight modules that allow parameter-efficient fine-tuning of pretrained models. Specialized language and task adapters have recently been proposed to facilitate cross-lingual transfer of multilingual pretrained models (Pfeiffer et
al., 2020b). However, this approach requires training a separate language adapter for every language one wishes to support, which can be impractical for languages with limited data. An intuitive solution is to use a related language adapter for the new language variety, but we observe that this solution can lead to sub-optimal performance. In this paper, we aim to improve the robustness of language adapters to uncovered languages without training new adapters. We find that ensembling multiple existing language adapters makes the fine-tuned model significantly more robust to other language varieties not included in these adapters. Building upon this observation, we propose Entropy Minimized Ensemble of Adapters (EMEA), a method that optimizes the ensemble weights of the pretrained language adapters for each test sentence by minimizing the entropy of its predictions. Experiments on three diverse groups of language varieties show that our method leads to significant improvements on both named entity recognition and part-of-speech tagging across all languages.
Language models are notoriously difficult to evaluate. We release SuperSim, a large-scale similarity and relatedness test set for Swedish built with expert human judgements. The test set is composed of 1,360 word-pairs independently judged for both r
elatedness and similarity by five annotators. We evaluate three different models (Word2Vec, fastText, and GloVe) trained on two separate Swedish datasets, namely the Swedish Gigaword corpus and a Swedish Wikipedia dump, to provide a baseline for future comparison. We will release the fully annotated test set, code, models, and data.
يعرف الموظف العام بأنه كل من يعمل في خدمة مرفق عام يديره أحد الأشخاص الاعتبارية العامة, و بشكل دائم, و لا يدخل إلى الوظيفة العامة إلا من تتوافر فيه الشروط التي حددها القانون , مثل الجنسية و سلامة الصحيفة العدلية و المؤهل العلمي, و السن, الخ...
Exam ) objective tests) have an important role in the evaluation of student's intended
learning outcomes, and estimates the level of their meeting of the sought goals. Therefore,
the skill of formulating exam tests is one of the main criteria that
should be within criteria
of the evaluation quality. An objective test is one of the sorts that offers a student with a
number of answers to a question or solutions to a problem, and then the student should
identify (recognize) the right answer or solution to these questions. They are called
"Objective Tests" because they are objectively corrected, in other meaning, its correction
doesn't differ from one evaluator to another. Because the importance of formulating this
type of questions and scarcity of researches concerned this idea. This research aimed to
evaluate the achievement of the formulated objective tests for formulated objective test
conditions. The taken research sample was the final written exam (second term) which
represented the courses of the four years that are taught by faculty staff members at
Nursing Faculty for the academic year 2014 -2015. The sample was consisted of 719
multiple-choice question, 248 True/False item, and 21 Match item. The most significant
findings were: the faculty members followed the conditions of formulating stem in
multiple-choice questions with the need to pay attention to language clarity and accuracy.
In addition, they followed the conditions in formulating alternatives in multiple-choice
tests, and follow all test-type formulation conditions for True/False items, with relying
heavily on multiple-choice questions and True/False items, while relying on the test of
match items (proportionality) type was very few.
The research aims to estimate the effect of sample size on the statistical test
power (t) for one sample, two interrelated samples, two independent samples,
and on the statistical test power of one-way analysis of variance test (F) to
compare the
averages. The descriptive method was used, and different sizes
of samples (300) items, where it was generated using the program (PASS 14),
and taken into account to be realized in this data the set of assumptions
needed to make test (t) and (F), with respect to random testing, categorical
level of measurement, normal distribution, and equinoctial variance.
This study aims at determining of cut-off score of the criterionreferenced
test that measures the fifth primary grades ability to solve
verbal problems that require carrying out division and other calculations
in homs by using contrasting groups m
ethod and making sure of it
efficiency of Lessing from the classification errors comparative with the
sample mean with the volume /400/ male and female pupil.
The study concluded that the classification errors yielded by the
cut-off score identified by contrasting groups method /74/ were less than
the sample mean /66/ of which to point out efficiency the used method at
this study.
The research aims to show the importance of the application of
International Standard ISO 17025: 2005 to adjust the general
requirements of the test laboratories efficiency. The importance of
this international standard by giving great importance
to the test
methods and calibration, which stressed the need for the
documented methods to carry out tests and analyzes and operation
of devices, and that it leads to work systematically and preserve the
expertise & documenting them and discover the analytical
problems and resolve them.
The efforts of the software developer teams are focused on conducting tests to detect different types of errors in a systematic way, with the least amount of cost, time and effort.
