-
-
When run
hidden_states, residual = fused_add_norm_fn(
hidden_states if residual is None else self.drop_path(hidden_states),
self.norm.weight,
self.…
-
Hi,
I want to try to change the backbone network into MobileNet-v2 with FPN. Is there any suggestions? THX!!!
-
### Description
I am trying to use inspec to test whether Remote Desktop Services on a Windows Server 2012 R2 box have been correctly configured. As a first test I want to make sure that the RDS Se…
-
Hi Dr.Lu
I found a discussion of adaptive activation functions in the open issues, and you mentioned that the "_https://deepxde.readthedocs.io/en/latest/modules/deepxde.maps.html#deepxde.maps.activat…
-
During versi burn-in of https://github.com/paritytech/polkadot-sdk/pull/4874, rev. https://github.com/paritytech/polkadot-sdk/pull/4874/commits/8f737b3de5802d9a74ff7006860ef1d696c6bac3, approximately …
-
Base on tiny-dnn, and refer to darknet and caffe, I rewrite a light-CNN which only uses CPU.
The following are some experimental results (I use a i7-6700 4cores 8threads, 8GB desktop).
1. MNIST, ach…
-
Anyone has interest to utilize the sparsity to accelerate DNNs?
I am working on the fork https://github.com/wenwei202/caffe/tree/scnn and currently, on average, achieve ~5x CPU and ~3x GPU layer-wi…
-
I'm using Leela Zero 0.17 + AutoGTP v18.
With one gpu it works fine.
But with two gpus it doesn't work.
Lizzie 0.6 is open but:
Leela Zero is loading...!!!
I can use the x button and do moves o…
-
Since we can handle generic graph-networks now, it's time to implement LSTM/RNN. Milestones:
- [ ] Add a basic LSTM cell (equivalent to TF's BasicLSTMCell)
- [ ] Modify network class to handle sequent…
nyanp updated
7 years ago