-
**I write a slowfast_custom_config.py . It reads:**
model = dict(
type='Recognizer3D',
backbone=dict(
type='ResNet3dSlowFast',
pretrained=None,
resample_rate=8,…
-
I tried distributed training slowfast model, unfortunatey i can not run it sucessfully.
I use two nodes to running the training:
the scripts are:
python tools/run_net.py \
--cfg path to config…
-
Excuse me, how to solve this problem？
Traceback (most recent call last):
File "inference.py", line 94, in
s_latent_a = static_enc_a(static_a)
File "/opt/conda/lib/python3.7/site-package…
-
I‘ve found that the Slowfast used the SGDR and the base LR was set to 1.6 in K-400.
how does the base LR be settled?
-
Dear SlowFast-Team,
this line of code results into an error, because the AVAMeter-class has no attribute video_preds.
https://github.com/facebookresearch/SlowFast/blob/cec2d0a3f17e5d152809899930…
-
## 🐛 Bug
I don't know if it's a bug or not.
I am trying to convert my slowfast torch model to onnx using `torch.onnx._export`. When I converted it to onnx and visualize that i got to know there is u…
-
I am very interested in X3D: Expanding Architectures for Efficient Video Recognition paper and wonder when will its implementation be available in SlowFast repository. Is it in progress or is there an…
-
Hi.
I want to use a model to classify 1 class. I use pre-trained weights for SLOWFAST_8x8_R50 to do this but get various errors. Ultimately, I get an error from the metrics script in the ks list. Ho…
-
Hi, I am trying to test the model-zoo model: TIN trained on MMit dataset based on resnet 50. I changed the default.yaml in ./experiments/tin folder to the following:
```
version: 1.0
config:
gp…
-
When running multigrid training, the code segment below is not compatible.
https://github.com/facebookresearch/SlowFast/blob/133e40f8349ce37b0e6168639da0811a413579c8/slowfast/utils/misc.py#L158-L161
…