-
As the 2010’s draw to a close, it’s worth taking a look back at the monumental progress that has been made in Deep Learning in this decade.
Tags: deep_learning
via Pocket https://ift.tt/…
-
## 概要
我在```https://github.com/okankop/Efficient-3DCNNs```中根据其论文方法去训练了一个模型。其模型是将传统的2D CNN模型```mobilenet resnet```等修改为可处理视频的网络模型。计划部署至Android平台作为手势识别。
## 问题
1. 其输入格式为```(N, C, F, H, W)```其F代表帧数。那么我该如…
-
how to add attention mechanism?
-
Hi there,
I like your GitRepo, and I have started to use it on my project.
I have a problem with the Embedding layer. If the first layer is the Embedding layer, the "see-rnn -> get_gradients" brok…
-
## 🐛 Bug
I have a ResNt18 which uses PReLU in some of its modules, therefore in order to quantize it, I had to rewrite the PReLU myself.
This is what I have comeup with :
```python
class PReLU…
-
Choosing an object detection and tracking approach for an application nowadays might become overwhelming. This article is an endeavor to summarize the best methods and trends in these essential topics…
-
hello everyone:
I am trying to implement asr and tts joint training by espnet1. When I use defualt libritts config under tacotron 2, the attention map is seems to be right. However, asr mod…
-
Mish:
𝑓(𝑥)=𝑥⋅𝑡𝑎𝑛ℎ(𝑠𝑜𝑓𝑡𝑝𝑙𝑢𝑠(𝑥))=𝑥⋅𝑡𝑎𝑛ℎ(ln(1+𝑒𝑥))
https://arxiv.org/abs/1908.08681
https://github.com/digantamisra98/Mish
-
this umbrella issue tracks my current progress and discuss priority of planned TODOs. It has been closed since all objectives are hit.
# Goal
- [x] achieve a high quality human-like text to spe…
-
hello.
Please help me, I do not know how to solve my problem problem.
I run and completed without errors
`python synthesizer_preprocess_audio.py `
`python synthesizer_preprocess_embeds.py /SV2T…