Do you want to publish a course? Click here

Feed-forward layers constitute two-thirds of a transformer model's parameters, yet their role in the network remains under-explored. We show that feed-forward layers in transformer-based language models operate as key-value memories, where each key c orrelates with textual patterns in the training examples, and each value induces a distribution over the output vocabulary. Our experiments show that the learned patterns are human-interpretable, and that lower layers tend to capture shallow patterns, while upper layers learn more semantic ones. The values complement the keys' input patterns by inducing output distributions that concentrate probability mass on tokens likely to appear immediately after each pattern, particularly in the upper layers. Finally, we demonstrate that the output of a feed-forward layer is a composition of its memories, which is subsequently refined throughout the model's layers via residual connections to produce the final output distribution.
The search was conducted (newborn feeding "premature" in the first week of life) in order to assess nutrition in newborn and learn about nutrition during the first week of life and what is the best food in preterm infant valiant hospital generously almonds-Pediatrics-premature Division.
This research presents a new methodology for the development of a controller based on Artificial Neural Networks and Direct control method in order to obtain the maximum available energy from Solar Photovoltaic (PV) Energy systems under different a tmospheric changes of the solar insolation and ambient temperature. In this context, this research presents a new model for MPPT-ANN in order to track the Maximum Power Point of PV systems in Matlab/Simulink environment. The developed controller is based on Feed Forward Neural Network FFNN trained by Back-propagation algorithm of error to determine the optimal voltage operation of the system PV system at different atmospheric changes. This research also suggests, control algorithm based on the direct control method in order to determine the duty cycle, which used to control directly the operating of DCDC Voltage Converter, depending on a comparison of the difference between the output voltage of PV system and the optimal voltage output of the neural network. The developed controller MPPT-ANN based on a network FFNN, Characterized by fast speed to track of MPP point and achieve high efficiency for the PV system under the atmospheric changes. The simulation results completed in Matlab/Simulink environment, showed the best performance of developed controller MPPT-ANN by achieving a better dynamic performance and high accuracy when tracking the MPP, compared with the use of the another PI-ANN controller based on artificial neural network and the conventional Proportional-Integral Controller, and compared with the use of the conventional MPPTP& O based on Perturb and Observe (P&O) technique under different atmospheric changes.
A prospective study included eighty tow patients (57 male) the range ages were (42 year) underwent surgery and developed postoperative enterocutaneous fistulas in the department of general surgery in Alasad and Almuasat Universitiys Hospitals betw een 2014 and 2016 .After the diagnosis performed ,They were approached according to specific strategy, and determined the onset of developing the enterocutaneous fistulas from the surgical procedure. the rate of developing in the first week was 68.2%. Accompanied symptoms and resulted complications were studied and classified according to an essential criteria.
Macrolophus caliginosus (Hemiptera: Miridae) is a zoophytophagous predator widely used in integrated pest management programs in both greenhouse and open-field . Mass rearing of Macrolophus caliginosus is greatly dependent on Ephestia kuehniella Z eller (Lepidoptera: Pyralidae) eggs as food source. Moreover, the addition of this factitious prey after the inoculative releases of Macrolophus caliginosus under field conditions is recommended to facilitate establishment of this mirid. However, E. kuehniella eggs are expensive and availability is limited. One possible strategy to reduce the amount of E. kuehniella eggs needed is the provision of sugar.This study was conducted during the period between May and November in 2014 in the labs of the Lattakia center of rearing bio enemis. In this work, the effect of sucrose as nutritional supplement on selected life-history traits of M.caligenosus was studied. The addition of sucrose (100 g/l) ad libitum to a diet of E. kuehniella eggs significantly increased the progeny of Macrolophus caliginosus and did not affect survival of nymphs nor developmental time. Moreover, addition of sucrose significantly reduced the number of E. kuehniella eggs consumed. These results may have practical implications of interest in mass rearing systems of M.caliginosus and its management in fields and greenhouses as a part of integrated control programs.
The experiment was carried out at the end of April during 2013 In Abi- Jarash farm, Faculty of Agriculture, Damascus University in order to determine the influence of organic fertilization on the growth and yield of Snap bean. The snap bean ”Tema” Phaseolus Varity, and organic fertilizer” Humax” with two methods (root application and foliar spry) and two concentrations (2 and 4 g/L) were used. The organic fertilizer was used after two weeks of germination and then two weeks between treatment and another and the experiment was laid according to simple randomized block design. Results showed that organic fertilizer caused an increase of plants growth and yield, and the foliar spray method showed significant superiority (p<0.05) to some of the studied parameter such as plant dry weight (6.89 g), leaf area (4595.44 cm2), leaf area index (2.55), number of clusters (15.3 cluster/ plant), number of pods (29.6 pod/plant), plant yield (156.9 g/plant), compared to the root applied method. Results also showed no significant effect for the fertilizer concentration.
Field experiment was conducted during the season (2011 – 2012) at Bhanine village in Tartous city. Maize seeds of Ghouta 82 variety were planted in plastic pots filled with calcareous soil .Treatments including three levels of B (0, 3, and 6 kg.ha− 1) and four levels of Zn (0, 8, 16 and 24 kg.ha−1) added to the soil, in a completely randomized block design with three replications, to study the effect of different levels of boron and zinc and their interactions on maize content of some nutrients. Results of this study showed that there was a clear effect of B and Zn and their interactions in soil on leaf content of Mn, Fe and Cu. High amounts of zinc in the soil had a clear effect on reduced leaves content of B and increased its content of K. The presence of a of B in the soil, assisted to increasing of leaf N content. There was a significant effect of B and Zn interaction on leaf content of Zn, and no effect on leaf P content.
أجري هذه البحث بهدف معرفة تاثير إضافة الأحماض الدبالية مثل حمضا ( الهيوميم والفولفيك ) في الخلطات العلفية للفروج في الكفاءة الإنتاجية
The study was carried out at the experimental station, Faculty of Agriculture. Damascus University. 1464 of the hybrid Lohman broiler chicks, one day old, were divided randomly and equally into 8 groups (182 chicks/ G) with 3 replicates of 61 chic ks each. Chicks in each replicate were reared separately on deep litter in an independent room of open housing till the age of 42 days. All housing and management procedures were similar for all chicks while the feeding was different throughout the fattening period in terms of the amounts of lysine and methionine added to the mix feed. Chicks of G1, G2, G3 and G4 were fed on plants mixtures supplemented with 1, 1.5, 2 and 2.5 kg/ ton of methionine, respectively. Chicks in G5, G6, G7 and G8 were fed on the same mixtures of the previous groups but supplemented additionally with 1kg/ton of lysine. Result showed that adding lysine 1kg/ton and methionine 2kg/ton to plant mixtures of feeds of broiler helped in improving significantly most of the indicators of productivity and this in turn led to a reduction of the cost of feeding and to produce 1 kg live weight, thereby raising the index profit from the process of fattening and increasing the gain from fattening process.
When estimating groundwater recharge in karstified and fractured aquifers it’s important to take into consideration, that they differ from porous aquifers by inhomogeneities of their hydraulic properties because of the great variance of infiltrati on properties in their rocks. In this research the process of groundwater recharge in Sher-mansour and Haramon areas in Antilebanon, where karstified and fractured aquifers are developed, was discussed. The portion of rainfall reaching the groundwater was calculated by applying water balance and groundwater level changes methods. In the first method the area was divided into surface zones and water balance components were calculated in each zone and the portion was estimated at 55.8% of rainfall in dry years like 2005-2006. In the second method the area was divided into subsurface zones and groundwater level fluctuations, storage coefficient and groundwater storage were calculated. The groundwater recharge varied between 17.5% in Haramon area and 80% in Sher Mansour area. The results showed that portions of rainfall reaching groundwater are not equal in the deferent subsurface zones and are deferent from those calculated by water balance method.
mircosoft-partner

هل ترغب بارسال اشعارات عن اخر التحديثات في شمرا-اكاديميا