-
### Summary
Per the resources that @agahkarakuzu kindly shared with us in https://github.com/ME-ICA/tedana/issues/464#issuecomment-583893763, T2* estimates should be corrected for background field gr…
tsalo updated
1 month ago
-
Hello, I was trying to profile Megatron-LM (which is in example) and I got an error.
My enviroment is
Nvidia NGC Container pytorch 22.02-py3
pytorch 1.11.0
CUDA 11.6.0
Ubuntu 20.04
…
-
After we finalize the implementation for a homogeneous semi-infinite substrate (basic surface mode) - #101, we need to think about extending it to multi-layered substrates. Here are several issues wit…
-
Hello, sorry, I am sorry to disturb you. I hope that you can provide the frame tag json file of Penn-Action, because the task of re-labeling the data set is really huge. Maybe you can share some tfrec…
-
I was able to successfully train multimer on a singe node gpu with multiple gpus, but I have been having trouble modifying the training example to train on multiple nodes. Would it be possible to prov…
-
有三个问题请教:
(1)有关多任务梯度更新的,我对代码理解是这样的:(multi_task/run.py)
task1: 产生loss1,更新一次模型参数
task2:产生loss2,在上一次梯度更新基础上再更新一次
不断循环上述两个过程
(2)下面粘贴train.py中build_backward函数部分代码
我理解 param_list中就是存放模型参数值,那么updated_…
-
Hi, thank you for sharing your code!
But, I can not run it on GPU.
The cmd line like python main.py --device 'cuda:0',
and the Traceback message likes:
Training configs: Namespace(batch_size=32, …
-
Hi guys,
I've been reading benchmark_cnn.py and trying to figure out what batch_size really is in a multi-GPU setup when using it to calculate num_batches_per_epoch and decay_steps.
num_batches_…
-
**Describe the bug**
When running optimized whisper medium int8 on macOS coreml I get the error
```console
CreateSession(Msg("Could not find an implementation for DecoderMaskedMultiHeadAttention(1)…
-
hi
Do you know where I could get more info on the unet implementation ?
for example it says 'We expect 572 x 572 x 1'
it would be great to know why, and also what was the best accuracy, which hype…