-
https://whatasmallship.github.io/2024/06/14/Principles-of-Large-Scale-hhMachine-Learning-Lecture-3/
Lecture 3 Exponential Concentration Inequalities and ERM 回顾:切比雪夫不等式 0−10-10−1经验风险: R(h)=1n∑i=1…
-
## 一言でいうと
BigGANを利用して表現学習を行う研究。実サンプルx/Encoderで生成した潜在表現zと、Generatorから生成したサンプルG(z)/潜在表現分布からサンプルしたzの距離を近づける形で学習を行う。この形式の学習が、潜在表現的にも画像生成的にも有効であることを確認。
![image](https://user-images.githubusercontent.…
-
Dear Fluid TOC members and maintainers,
I am proposing to have Shiming Wu(GithubID: wushiming540) as a new committer.
Wu Shiming, senior engineer of Huawei's NAIE platform, is responsible for l…
-
-
## 一言でいうと
ロボット操作に強化学習を使う際、学習時間が大きな課題となる。そこで、7種類のロボットで記録した1500万のビデオフレームのデータセットを公開。様々なロボットの軌跡データから表現学習(次フレーム予測/過去フレーム予測)を行うことで、Zero/Few shotが可能な転移性能の高いモデルを構築できる。
### 論文リンク
https://arxiv.org/abs…
-
作者您好,我对你的研究非常感兴趣。不过我有一个问题,数据集中的npy文件是怎么得到的呢,是通过’data/image/common‘中的数据来进行变换的吗,还有seq_mean.npz,是通过什么得来的呢?请问作者有没有相关的转换脚本可以提供呢?谢谢
-
- [ ] [[2204.02311] PaLM: Scaling Language Modeling with Pathways](https://arxiv.org/abs/2204.02311)
# [PaLM: Scaling Language Modeling with Pathways](https://arxiv.org/abs/2204.02311)
## Snippet
"…
-
https://whatasmallship.github.io/2024/06/17/Principles-of-Large-Scale-hhMachine-Learning-Lecture-4/
Lecture 4 Learning with Gradient Descent 回顾:经验风险最小化与梯度下降 为每个预测器分配一个ddd维的参数向量,也即每一个ddd维参数向量对应一个预测…
-
To validate and measure the effectiveness of recent improvements in the learning process, a comprehensive large-scale simulation run is necessary. This will help ensure that the enhancements not only …
-
https://www2.eecs.berkeley.edu/Pubs/TechRpts/2017/EECS-2017-219.pdf