-
-
runninnggggngnnngngg yayyyy
-
hi @qqwweee
Conv-lstm has been added to Darknet and I am trying to convert trained yolov3+conv-lstm to .h5 using your repo (convert.py ).However, it has unsupported header error .I really need to …
-
you are basically using stacked lstm architecture right?
-
## 一言でいうと
LSTMを強化する前処理の提案。入力/隠れ層を双方向にgateしていくことでcontextualizeし、表現力の高まった入力/隠れ層をLSTMに入力する。単純な処理ながら言語モデルで良好な精度(perplexity)を確認。
![image](https://user-images.githubusercontent.com/544269/72582651-c41…
-
-
## 一言でいうと
LSTMにおいて層を重ねるのでなくネストした方が良いという研究。具体的には、LSTMのcontextを更新する代わりに内部のLSTMへのinputとして送り込む(外側のforgetとinputからの出力は通常合算でcontextの更新に使われるが、これをコンカチし内部のinputに送る)。PTB、特に中国詩データセットで顕著な効果
### 論文リンク
http…
-
**Describe the Question**
Hello,
I constructed Stacked-LSTM in CustomPolicy module and here’s the point that Stacked-LSTM may perform better than Single LSTM reasonably but we got the exactly opposi…
-
Could you please share the features for the LSTM inputs? I am trying to reproduce them but I cannot reach 10% of improvement only by using LSTM as you report in the paper. For instance, I cannot impro…
-
Would this also work for LSTM? Do you have any benchmark of the method on that?