Stone comments

Results 17 comments of


                                            Stone

About thread mode

Usually, events may occur in many places. The response events need to be processed on the same thread. Among the currently provided threadmode, only main has this characteristic, and other...

Issue for dwsHashtables.pas

The line `for I := #1 to #255 do` may be more than 255 times cycles. The I was type Char (WideChar). This value is likely to have implicit conversion...

Parallel.ForEach<> inside Parallel.Future<> seems to leak threads

See issue #194. The memory is not leaking, but it seems that the memory has not been released immediately, and repeated calls can lead to memory depletion

The training dataset is extremely large - how should it be handled in CAI?

I'm using this exact approach right now. But when it runs for an extended period, can we save the training state and pick up where we left off later?

The training dataset is extremely large - how should it be handled in CAI?

Thanks.

May I build an RNN or LSTM network?

Thanks

Can a 1×n convolution kernel be applied to a row vector

That's great

Can a 1×n convolution kernel be applied to a row vector

For example: I want to try applying a convolutional neural network to the input. The input contains such dimensional information : x = 100 y = 1 depth = 2....

Can a 1×n convolution kernel be applied to a row vector

Thanks

Can a 1×n convolution kernel be applied to a row vector

@joaopauloschuler I don't know if my thinking is correct. When performing convolution on tensors with y equal to 1, if padding is required for the convolution kernel, it can only...