-
## 一言でいうと
DNNに組み込める巨大なメモリ機構の提案。Key/Value型だが普通に実装すると計算コストがかかるため、Keyの作り方を工夫している。Key空間を2つのsub-key空間に分割し、クエリも2つに分割、各分割クエリに対応するsub-key空間で最も近いsub-keyを取得し、その外積でキーを作成する。
![image](https://user-images.git…
-
## 一言でいうと
多言語対応の文表現を得る際、どんなタスクが良いのか検証した研究。ベースは言語モデルで、通常通り次の単語を予測する(Causal LM)、単語をdropした箇所を予測する(Masked LM)+翻訳データがある場合に、並べた文でMasked LMを行うTranslation LMの計3つを提案。CLM
-
Dear Dardouri,
I wonder if you could provide some hints on how to do inference on your model, i.e encoding and decoding? Which part of your code should I use?
We are the group from University of…
-
1. Provide comments to the code of conduct
2. Dissemination : define a timesheet for social media management (semanero)
3. Webinar oganization
4. Report Sistemas complejos
5. Draft doc…
-
Demande d'un utilisateur :
"Je souhaite qu'il y ait un petit guide d'utilisation du site web : Le FAQ pourrait être utile aux nouveaux utilisateurs. Là, nous pouvons expliquer nos nouveaux points fo…
-
## 一言でいうと
DNNの実行で、再利用可能なモジュールを組み合わせる試み。全体としては強化学習(AlphaZero)の枠組みで、環境以外にプログラムのembeddingを受け取りLSTMで処理、Action/Valueを出力するという流れになっている。LSTMはプログラムの実行環境として動作し、STOPというactionが出るまで実行を続ける。
![image](https://u…
-
Hi,
I am trying to extract features for coco dir - however it always gives me RuntimeError: CUDA out of memory.
I have tried using single GPU by setting CUDA_VISIBLE_DEVICES=0- still ti gives me t…
-
```
import pandas as pd
# Assuming your dataframe is named df
# Example of dataframe creation
# df = pd.DataFrame({'matched_name': [...]})
# Dictionary of university names to replace
universities =…
-
Hi,
I'm trying to use DiAna, but as soon as I open the DiAna_segment, I have the following error.
Can you help me to figure out how to solve it?
Thank you.
![image](https://user-images.githu…
-
Revising ones that clearly shouldn't be organizations ex. Evangelicalism.
Revising standard names that are acronyms when they should be full forms, possible duplicates that have arised from this. ex.…