Detectron.pytorch

Detectron.pytorch copied to clipboard

python tools/download_imagenet_weights.py

2

Traceback (most recent call last): File "/home/zq/anaconda3/lib/python3.6/site-packages/urllib3/connection.py", line 141, in _new_conn (self.host, self.port), self.timeout, **extra_kw) File "/home/zq/anaconda3/lib/python3.6/site-packages/urllib3/util/connection.py", line 83, in create_connection raise err File "/home/zq/anaconda3/lib/python3.6/site-packages/urllib3/util/connection.py", line 73, in create_connection sock.connect(sa)...

zq130320339

updated spelling of the default CUDA_PATH in readme

arvind-iyer

Combine train and val?

8

Hi, just ask, would you consider to combine the train and val together? Like a standard one: every epoch do the validation before ckpt, if val accuracy/loss is higher then...

QiaoranC

Hi, `torchvision 0.3` has been released and supports ops such as `nms`, `roialign`. It is now possible to support PyTorch 1.0.

inrainbws

Inference on CPU

1

Is it possible to run inference on CPU? In the forward function of the roi_Xconv1fc_gn_head_panet in fast_rcnn_heads.py, it relies on the gpu version of the roi align. How can this...

ashnair1

update README.md , add compiler version note

1

add compiler version note according to https://pytorch.org/tutorials/advanced/cpp_extension.html#using-your-extension

timfaner

* fix typos on strings 75 & 277 * syntax issue on string 144

kant

I need to test new backbone of faster-rcnn, it it a NAS model. Can you tell me how do I add new backone?

xiaozhuka

training process acquired abnormal weights, i.e., nan

1

## PLEASE FOLLOW THESE INSTRUCTIONS BEFORE POSTING 1. **Read the README.md thoroughly ! README.md is not a decoration.** 2. Please search existing *open and closed* issues in case your issue...

19940312

loss_bbox almost equal to 0

3

I attempt to run the faster rcnn fpn with resnet50 model on my own dataset. But I meet that the loss_bbox is almost to be zero and the final mAP...

liuliu66